4.16.2010

Intel and AMD Report Record 1st Quarter Revenues

What a difference a year makes. The start of 2010 ended what can only be considered a disastrous 2009, not just for semi-conductors, but for the market as a whole as we begin to recover from the worst recession in over 80 years. It is clear that Uncle Ben Bernanke's 0% interest rates are fueling the next great bubble to replace the real estate bubble which Greenspan induced as a result of the Dot-com bubble. Where this bubble will go is anyone's guess (My bet is on T-Bills and the Federal Government's record deficits.) Meaning we are experiencing a nominal recovery of increased spending, as opposed to a real recovery on sound fundamentals, but I digress. Regardless of the cause of the increased demand, businesses are doing everything they can to soak up as much of it as they can and Intel and AMD are no exception.

Both companies reported record 1st Quarter revenues. Coming off such a dreadful 2009 makes this even more impressive. Whether this is a result of a strengthening technology sector, the release of new products or the unleashing of pent up demand from consumers and businesses from last year (perhaps a combination of all), this was a great quarter for both companies no matter how you slice it.

INTC


Great numbers and results all around for Intel. Gross margins continue to improve to mind boggling levels. At this rate Intel will post the largest yearly profit in history. A few things that stand out from the report, strong growth in nearly every sector except for Atoms which were down significantly. Despite Intel's assurances that Atoms and Notebooks did not cannibalize each others sales, I can't help but see a correlation with last quarters drop in Atom sales with that of record mobile chip shipments. Whether this was "good" or not is hard to tell since both products are profitable, but it's a great hedge strategy for all market conditions. Outside of normal growth and expansion, the Atom's make for a great value segment in down markets and Notebooks are the perfect product for booms. Either way, Intel can take market share.

We also saw the release of Nehalem-EX for the MP server segment which replaces the much maligned Dunnington platform for a true competitor in this space to take on AMD in this segment in a way that has been lacking for some time. It may not have made a large impact on financials due to its late release in the quarter, but chips were shipped for revenue. In the mobile space, we saw the release of the Westmere family chip Clarkdale which likely drove the strong increase in mobile sales. Also, the conference call noted that OEM inventories were normal but Intel inventories were lean. Indicating the upside of last quarter was above Intel's internal estimations.

AMD


Gross margins are up to a healthy 47%, operational profits are reported at a strong $182M and a net profit of $257M. Considering this was a "seasonally down" Q1 makes this even more impressive. The only caveat is the inclusion of a $325M non-cash assessment due to "deconsolidation" with Global Foundries. Excluding this one-time gain will bring AMD's net profit to a narrow loss for the quarter. Which means they essentially broke even. If the market continues to improve, it's likely that AMD could even report a net profit for the calendar year in spite of one-time accounting tricks. Not only were CPU shipments strong, but GPU shipments increased sequentially due to a strong ATI lineup.

AMD didn't sit idly by with old products going into the new year. We saw the much anticipated release of Magny-Cours, the MCM 12-core server chip (2 x 6 core) to compete in the EP space which Gulftown had taken by storm. We also so a flurry of higher binned Phenom II's giving AMD a strong lineup in the mainstream segment. Overall, a very good quarter with the signs of finally breaking out of the continual slump of heavy losses.

197 comments:

Anonymous said...

I hate how AMD still does these games as surely next year when they are doing year on year comparisons they'll be very quick to omit the 325mil to make that comparison look better. Though even with the 325mil charge taken out the numbers are not bad. It's hard to draw comparisons to the street estimates when these things are thrown in...

The best comparison seems to be the revenue of 1.57Bil vs a consensus of 1.5Bil - its a beat but it's starting to get into noise range. On paper it's a 5% beat but I doubt analysts are going to a 2nd decimal place in their estimates.

The non-GAAP margin was 43% and is in line with AMD's projection, the 47% # is once again smoke and mirrors and makes for a better sounding # (but isn't real).

Anonymous said...

At the previous poster, AMD does include the cost of making its processors in GLOBALFOUNDRIES. I did not surprise what we have to expect from AMD's financial report.

SPARKS said...

Korgano, excellent report, well structured, accurate, and unbiased; well done.

Your observation on the ATI graphics component is particularly important. The 5xxx series success has been nothing short of brilliant. Couple this with NVDA’s incessant delays, compounded with extremely low yields plus TMSC’s failure to deliver 40nM has been a recipe for a brilliant AMD recovery and success.

Even the G himself didn’t see that one coming. (I believe he’s a bit hard on himself. See his last comment on the previous post. He still made a 50% profit on AMD’s wild stock fluctuations. A Ballsy move, and I said so then and now. ) In fact, who the hell could have seen the INTC settlement, NVDA’s flop, TMSC’s flop, the GF divestiture, and the wildly successful 5870 card all happen in one years’ time!?!? Only a deeply entrenched insider, that’s who.

As far as what the future holds for AMD will be based on process implementation. They could design revolutionary designs on paper without the burden of a foundry model from now till the four horsemen come galloping down 5th Ave. But, who’s gonna lay ‘em down successfully on silicon past 40nM?

I’m quite certain there’s one company is cooking up 22nM wafers as we speak.

SPARKS

Anonymous said...

At the previous poster, AMD does include the cost of making its processors in GLOBALFOUNDRIES. I did not surprise what we have to expect from AMD's financial report.

I'm not talking about the cost of wafers from Global foundries... the 325mil had nothing to do with buying wafers, it's an accounting # which they attributed to deconsolidating from GF. The expenses you are talking about are elsewhere in the financials.

Also of note, AMD took something like a 185mil charge (for losses due to their ownership stake in GF - which I guess was not profitable... shocker). Look for them to continue to dump their share in GF as quickly as possible (I think I remember seeing that they already lowered their stake) They also had some sort of interest gain....

Bottom line - they were +211mil in these one time benefits and charges. And again none of this has anything to do with buying wafers or R&D or normal operating expenses or revenue. The "real" net income was just under 50mil. This is more or less breakeven, and is a good result for AMD given their recent history, but let's not make this out for more than it is thanks to 211mil in one time benefits being lumped into the accounting. (just take a look at how the stockprices reacted post INTC earnings and post AMD earnings - while the street doesn't always get it right, they get it more often than not)

I'd expect AMD to perhaps go slightly back to a loss in Q2 and maybe be profitable again in H210.

A Nonny Moose said...

I noted that AMD's stock took a whack after some analysis of the Q1 results. Compared to Intel's record-setting Q1, it's clear AMD has to raise their ASPs sometime soon - don't they still have another 1~2 billion in senior notes due in 2013 or so?

IIRC until Intel started shipping i5's last fall, AMD was getting around $260 or so for the P2-965, which Newegg now has for $180, and the 6-core Thuban is now in that price range with 50% more die space. And it doesn't look like Magny Cours will do a lot in the server space either.

IMO AMD needs a big boost from Bulldozer and Llano or they are going to have a lot of trouble paying off the remaining notes. Unless of course they get another bailout from Intel :P.

Anonymous said...

Pretty sad results from AMD, with such a rising tide anything but the leakiest boats would do better than this sorry A$$ company

Anonymous said...

IMO AMD needs a big boost from Bulldozer and Llano or they are going to have a lot of trouble paying off the remaining notes. Unless of course they get another bailout from Intel :P.

AMD needs the industry to grow and needs to stop the obsession with marketshare via pricing strategy. I know I've beat this dead horse way too many times but with the rumored 6 core desktop chip pricing it seems they haven't learned...

The other avenue is notebook space where they are woefully behind in marketshare (I think it's in the low 10's). Maybe fusion will help, but again the majority of people don't buy notebooks for 3D gaming so if they try to take that marketing avenue it will fail.

It seems AMD has a favorable deal with GF on the foundry pricing (at least for some amount of time) but I don't think it's in their best interest to play the pricing game and just push the losses onto GF (and apparently hope GF's other customers pick up the slack). Without a healthy GF, AMD is done.

So they need to balance trying to pick up volume which is probably coming from GF at close to zero margin (maybe even at a loss for GF) and focus on making more per chip as opposed to selling more chips. Until GF can balance their capacity with many other customers to keep their fabs loaded, AMD taking up more capacity might actually be a bad thing for GF (which in turn is bad for AMD in the long run). In other words, in order for GF to grow capacity they need to be getting decent margins on existing capacity to fund future expansions (or rely on more suckers in NY). The more capacity on the AMD side, the lower the margins.

SPARKS said...

(or rely on more suckers in NY.)

Thanks G, you know how much I love you when I vomit.

SPARKS

SPARKS said...

Hey folks, AMD is going bulk. So much for SOI.

http://www.xbitlabs.com/news/cpu/display/
20100420225440_AMD_Ontario_Monolithic_System_on
_Chip_40nm_Fabrication_Process.html

Anonymous said...

Sparks - that's a different product... I'm pretty sure it has the Bobcat core (low power) in it as is more of an Atom type product.

Anonymous said...

Still waiting, where is the consortium's HighK Metal Gate. Its been years since they had their big press annocument. Where the hell is that cheaper more elegant gate first approch tat IBM/AMD/GF have been saying is so good. Intel has shipped hundreds of millions of HighK already. You'd think will all them reverse engineering they'd at least be able to copy it in two years!

I wouldn't be surprised at all if they go bulk, SOI is some more FUD, right up there with SILK and gate first I say!

Anonymous said...

nyah, nyah @previous poster.

AMD Magny Cours which have 2*700 mm^2 die are consuming 115 watt TDP in 45 nm process without need HKMG receipts. What if AMD has HKMG in 32 nm process for its processors, I think it's going to be interesting to see AMD is surpassing Intel in transistor switching performance.

Anonymous said...

I love AMD, I got big power hungry transistors, big money sucking die.

I go to NY state and arabs tor hand outs.

Damm who needs HighK metal Gate or small fast transistors

Maybe Obama will give me a charity hand out too!

Anonymous said...

Heh, you regret that Intel gave money to AMD. It was AMD's own fault that AMD did not have money to support its customers. Even, it's a no brainer AMD products are rubbish compared to Intel products because typically, AMD products are unreliable. These are the reasons why in 2003, Sony had dumped AMD Athlon completely in its VAIO lines and since then Sony has been using Intel processors exclusively.

SPARKS said...

"What if AMD has HKMG in 32 nm process for its processors,"

Huh?

IF they did, or rather, IF there were a leaking pissing hint there was, a couple of things would happen.

First, AMD would need to get back in the foundry business. Since they recently shed their stake in Global foundries, they only design chips; they don’t do process.

Secondly, IF Global foundries had HKMG at 45, forget 32, the news would be all over the web like Monica Lewinski cigars.

Third, IF ANYONE besides INTC had a HV HKMG process, like TSMC, you’d hear the fireworks from Hsinchu Science-Based Industrial Park, Taiwan in Ohaha, Nebraska.

Fourth, IBM (and the so called consortium) would be screaming from the roof tops.

Fifth, HKMG on SOI? Now that would be impressive.

And finally, I would be right behind the smartest person I’ve ever known, who just happens to be on this site, loading up on both TSMC AND AMD stock.

I think you need to change the tobacco in that dream pipe you’re smoking.

SPARKS

Anonymous said...

Well, tobacco is a good additive for a schizophrenic suffer like sparks.

SPARKS said...

"Well, tobacco is a good additive for a schizophrenic suffer like sparks."

Give that man a sheroot!

http://www.politicalgifts.com/monicacigars.html

Sparks

Anonymous said...

So is the clock ticking for INTEL

ARM is planning servers chips, should be cheap, scalable and power efficient.

ARM owns the smartphone, everyone that is smart will have ARM with flash. Pretty much everyone will have a smartphone very every person and one PC at home. I see 1 billion ARMs chips and 300 million x86.

INTEL should be very very scared. They really need to buy ARM or get their act togather and produce a power efficient x86 SOC. They got double the transistor budget and a far superior process where the hell are they?

What do you guys think, is INTEL going to be toasted by ARM and Samsung and TSMC just like they toasted AMD?

Anonymous said...

Or are they finally going to deliver something with their superior process and x86 and do to ARM what they did to all them RISC and AMD too!

nanikore said...

feenom ex six...

For the same price, Core i7 860 and Core i5 750 are generally better buys than Phenom II X6 1090T and 1055T.

The new Thubans excel at certain content creation tasks, and unless you're very specific about what you're going to do with those chips they're still not winners as far as performance per dollar is concerned.

Close, but no cigar.

Unknown said...

What do you guys reckon on AM3 vs LGA1366 and LGA1156, considering Sandy Bridge needs a new socket?

Unless you're like Sparks and require the absolute fastest thing at purchase time (which right now is Core i7 980X, for a million dollars), AM3 on 890FX sounds like a good choice. Stick a Phenom II X6 1090T in it for now, then drop a Bulldozer derivative in it some time next year..

http://www.semiaccurate.com/2010/04/15/intels-patsburg-replace-x58-2011/

A Nonny Moose said...

AMD needs the industry to grow and needs to stop the obsession with marketshare via pricing strategy. I know I've beat this dead horse way too many times but with the rumored 6 core desktop chip pricing it seems they haven't learned...

Yep - even that reknowned AMD fanbois Baron Matrix would agree with you on that one :P..

OTOH AMD probably prices their X6 where their market analysis says is the sweet spot, considering performance and other factors. According to one news report I read recently, world-wide CPU shipments increased something like 24% last quarter which is a pretty good rising tide. Too bad AMD apparently did not take advantage of it, according to their quarterly report. The world economy could easily tank once again - Israel bombing Iran's nuke plants and Iran shutting off a huge chunk of Mideast oil, for one thing. Without future Intel bailouts, AMD could go down for the count in another severe recession..

InTheKnow said...

AM3 on 890FX sounds like a good choice. Stick a Phenom II X6 1090T in it for now, then drop a Bulldozer derivative in it some time next year..

The problem I have with this statement is the implied assumption that Bulldozer is going to be a success.

I'm not a big follower of AMDs architectures, but it seems to me that the first iteration of opteron wasn't a world beater. The second iteration was the best processor available at the time. Phenom was a big disappointment. Phenom II is still behind Intel's products, but is what the first Phenom should have been.

Now you have another new architecture coming out. If past history is anything to go by, the first iteration of bulldozer won't live up to the expectations.

I could be wrong, and bulldozer could be the hottest new thing since sliced bread, but history would lead me to believe otherwise.

InTheKnow said...

I know I've beat this dead horse way too many times but with the rumored 6 core desktop chip pricing it seems they haven't learned...

No Guru, it is you who haven't learned. Intel needs 8 fabs to keep up with AMD. So we can clearly see that Intel's yields are rubbish while AMD, through GlobalFoundries, produces nothing but perfect wafers. Therefore, AMD can afford to practially give their chips away. :-P (sarcasm intended)

Unknown said...

You could be right ITK. However, upgrade to a next generation architecture (with the potential of being good), is better than no real upgrade path? Seems like Intel is prematurely moving on from LGA1366. It's a strong platform. Feels like both of them (LGA1156/1366) only came out yesterday, now they're being replaced (LGA1155/2011).

Anonymous said...

I'm always praying to the god because I wish Paul Otellini is becoming the US President. I want to see him bringing piece to the world. I believe if Intel has been successfully implanted its chip in everybody's brain, I believe American citizen will see their American dreams come true.

Anonymous said...

@InTheKnow, I believe everyone will be agree that Athlon was a game changer for AMD as an Intel competitor. Athlon had wins in the most benchmarks against Pentium III.

A Nonny Moose said...

http://channel.hexus.net/content/item.php?item=24418

AMD continues to lose CPU market share to Intel
Scott Bicheno - 29 Apr 10, 12:12pm

Market researcher IDC has released its figures for the Q1 global CPU market and they make good reading for the industry on the whole, with a 39 percent rise in shipments compared to the first quarter of 2009.

A look at the breakdown between the vendors reveals AMD is unlikely to be getting out the bubbly as it confirms that it has lost a fair bit of market share to Intel in that period.

In Q1 2009, AMD had 22.3 percent of the overall global PC CPU market, but a year later that has shrunk to 18.8 percent. While AMD has maintained a healthy 28 percent share of the desktop market, declines in both the mobile and server markets are what has led to the decline.

There was a decline in shipments from Q4 2009, but as Shane Rau, director of PC CPU research at IDC explained, that's to be expected. "PC processor shipments typically decline around 7 to 8 percent going from fourth quarter to first quarter. A decline of 5.6% is modest and wouldn't mean much by itself. However, after the huge rise in shipments we saw in the fourth quarter, it adds more credibility to market recovery and that the PC industry anticipates improvement in PC end demand in 2010."


From the server charts, it seems AMD did get a nice boost of 0.6 server CPU marketshare when Istanbul came out with no Intel counter, but that lasted only one quarter - probably the drop-in updates which fizzled out as the economy was still bad last fall. Now AMD has Magny Cours, but Intel also has the Westmere Xeons and the Becktons. I sorta think AMD's decline from 13.4% in 2008, to 10.1% in 2009, and to 9.8% last quarter, will continue, esp. since global CPU shipments increased almost 40% last quarter which should have floated everybody's boat, but AMD still either suffered a small loss or had a small profit depending on how you interpret their GF accounting trick.

Anonymous said...

Good stuff Anonnymoose...

It was fairly obvious from the respective revenue reports that Intel gained share, but it's nice to see how it broke down. It's got to be a bit scary that the one area where AMD is holding up (Desktop) is the market that's shrinking. And it also makes me scratch my head on the six core pricing - shrinking market... top end product (for them) which makes up an ultra small % of that shrinking market which pressures the pricing of the volume segment of that market... have I mentioned that before? ;)

but AMD still either suffered a small loss or had a small profit depending on how you interpret their GF accounting trick.

I think anyway you look at it they made a small amount of profit... AMD also had to include a one time loss from GF (I assume this was due to GF losing money for the quarter and AMD being a part owner)... if you sum up the one time benefit and the one time loss it was less than the net income they reported... it took them down to somewhere around +46mil or so... not as impressive as the $250mil they tried to trot out to Wall St but still profitable.

They guided revenue flat/slightly down so Q2 will be interesting as things could swing back to a loss... which might surprise a few folks considering they claimed a quarter Bil profit in Q1 (and I'm sure after the fact AMD will be quick to point out the one time adjustments made in Q1 so it doesn't seem like a large drop)

Anonymous said...

It's a good thing, Intel had decided to delay AMD's dead. Intel can steal another AMD's IP like Hypertransport external connection. So, if Intel had done in sucking AMD's IPs. Intel can flush the carcass of AMD in the toilet. My Dream when Intel will own all, is going to be a reality. I want Intel inside everyone and everything, because I think if Intel is the real ruler of the world, the world will be a better place for humanity.

Tonus said...

And thus we see the results of posting when you are high.

Anonymous said...

Sharikou and Scientia have been let out of their mommas basement. lets hope momma herd them back to AMDzone and their own fantasy land.

SPARKS said...

“What do you guys reckon on AM3 vs LGA1366 and LGA1156, considering Sandy Bridge needs a new socket?”

Hmmm, OK LEM, I’ll bite.

First, about Charlie at Semi Accurate, the name he’s chosen for his site speaks volumes. When Charlie gets his teeth into something substantial, he will pursue it with dogged persistence. His relentless, in depth analysis of NVDA has been nothing short of brilliant.

However, Charlie does have a penchant for feeding off the rumor mill and running with it. Sometimes he’ll simply throw something out there just to see what bites, while not strict journalism at its best, he has successfully hooked into quite a few big ones. But this report is FUD and everyone has jumped on it.

Secondly, this is old news and not a socket change, back in Feb. 2009.

http://news.softpedia.com/news/Intel-Socket-LGA-1155-1156-and-1567-Pictures-Leaked-103939.shtml

From everything I’m reading about the LGA 1155, from all accounts, the soon to be released 2011 socket was designed with the Integrated Graphics Controller in mind. It was planned for and designed to coexist with LGA 1156 years ago. The 1156 is far too successful and very deeply entrenched to simply abandon so early in the game.

The LGA 1567 is for the XEON beasties. And wonderful beasties they are.

Regarding the 1366, no way, I believe that socket will have the same long legs as the LGA 775, designed specifically for those wonderful performance nuts you spoke of earlier. There is no comparing LGA 1366 to anything else (except the 1156 of course). It’s simply no contest.

There may be comparisons with 1155 and 1156, perhaps even wins, but (as a performance nut) I couldn’t care less. I see these as lesser boards wrought with compromises to compete with you know who, performance per dollar, per watt, mom and pop, white box, budget box, bla, bla, whatever.

Hey, go with 1366 and a cheap i7, throw in a mild overclock and let them compete with you, simple. They’ve been chasing TONUS for over a year.

SPARKS

SPARKS said...

"There is no comparing LGA 1366 to anything else (except the 1156 of course). It’s simply no contest."

Whoops!


There is no comparing LGA 1366 to anything else (except the 1567 of course). It’s simply no contest.

SPARKS

SPARKS said...

G, Why, oh why, don't I ever listen to you? Well, I heard you, but I didn't do a damned thing about it.

http://channel.hexus.net/content/item.php?item=24440

SPARKS

InTheKnow said...

Well, I heard you, but I didn't do a damned thing about it.

And Intel has yet to be seen in this space.

I can't believe it bodes well that their 32nm Atom entry into this space isn't due to come out until 2012. They are missing their opportunity to get into the space early and will end up having to go up against firmly entrenched competition if they don't start getting some real design wins soon.

Having the LG smartphone pulled was very bad news in my opinion.

Anonymous said...

@InTheKnow, you don't need to worry if some Intel's customers have show their libelled behaviour to Intel. As long as Intel's research in brain implants has been showing a good progress, I believe it will not be an issue anymore. China government likes Intel's innovation because they want enable controlled behaviour of their citizens. I think many govenments in the world want to implement this technology to their citizen.

SPARKS said...

"against firmly entrenched competition if they don't start getting some real design wins soon."

Word. And it gets worse. Now that we have 3 APPLE devices in the house, my wife is throwing out questions about the iPAD.

BTW, 300,000 of the trendy clipboards sold upon initial release.

Love him or hate him, Jobs is marketing qenius.

ITK, regarding INTC missing the boat, you are starting to sound more like Moses than Chicken Little.



SPARKS

InTheKnow said...

ITK, regarding INTC missing the boat, you are starting to sound more like Moses than Chicken Little.

As long as it's not Noah. :)

Anonymous said...

Word. And it gets worse. Now that we have 3 APPLE devices in the house, my wife is throwing out questions about the iPAD.

By all means please purchase one... I own that stock too! I've never been a fan of their products, but they have a loyal following and have a brand recognition that allows them to charge a premium regardless of folks whining about cheaper products being available (reminds me a bit of the AMD fanboy price/performance whining.. some people want performance and are willing to pay for it)

The stock is probably over-valued compared to other stocks, but historically it has always run this way and every time I think about dumping it, they launch a new product which people will pan, but people will buy and I say 'let it ride'.

By the way AMD's down under $9... if it goes below 8 may be good for a short term trade (heck it might be good for a trade if it just gets in the low 8's). I think the industry overall will have a good H2. so if it's near 8 at the end of May or early June it may be worth a flier... (though you'd have to hope the street reacts to the Q2 report on outlook as opposed to earnings as I don't think Q2 is going to overwhelm anyone)

SPARKS said...

“If Chicken Little told you that the sky was falling,

Even if it was would ya come running back again?

I know you would my friend,

Again and again and again and again”

-Aerosmith

ITK, the call has been answered, ATOM Z600.

http://www.semiaccurate.com/2010/05/05/intel-announces-atom-z600-series-phones-tablets/

SPARKS

InTheKnow said...

Sparks, I believe this fits in with what I have been saying.

Moorestown is about 1/2 the size of the Menlow platform, so it is still on the largish size for a smartphone. If you look at the Moorestown smartphone prototypes out there like the LG phone, you will see that they are somewhat larger than the iPhone. Unless I'm very much mistaken this is due to Moorestown being a larger platform than that of the ARM competitors.

Idle power is still 2-3x of ARM but at 10 days of power at idle, I think that becomes more of a marketing point than a real issue.

A bigger concern is that the active power usage is still a bit high as well. From the article you linked to:

In terms of actual usage, things aren't quite as rosy, with a five hour battery life for smartphones and 10-12 hours for tablets. It's getting closer to the kind of battery life on offer with ARM processors, but it's still not quite there.

There is also the fact that the earliest these phones are projected to show up is the end of this year. That is another 6-7 months that ARM has to entrench themselves. Again from the article you linked to we have:

In the meantime ARM and its partners aren't going to sit around and idly let Intel steal customers. We should be starting to see dual-core ARM Cortex A9 based SoC powered smartphones and tablets in about the same timeframe as Intel is hoping to get Moorestown devices into the market.

And we have yet to see an actual product. So far all we've seen are prototypes at trade shows. Until there is something out there that you and I can buy Intel is just going to keep losing ground that they will have to fight tooth and nail to get back.

It is still my opinion that Intel needs the 32nm platform to be truly competitive. Unless Intel pulls it in, the schedules I've seen seem to be the 2012 time frame. Given the pace of innovation in this space, another 2 years seems to slow to me.

InTheKnow said...

Don't get me wrong, I think Moorestown will be an excellent product for tablets (which look like the next big thing as netbook sales flatten out). But I don't think it is really going to be an absolutely compelling solution to the smartphone space. To break that barrier I believe that Intel needs to get Atom on 32nm.

SPARKS said...

“Don't get me wrong, I think Moorestown will be an excellent product for tablets….”

Agreed, and per chance “Wintel” can produce a ‘PCpad’ that is as fully functional as a Netbook, i.e. word processing, hi-res photo’s, Excel, DVD play back, Wi-Fi, x86 based, decent speed, etc. ,they just may have a winner. (Anything but a damned APPLE) This would make Spock’s Tri-corder look silly.

Ah, yeoman Janice Rand, on a full 10 inches.

SPARKS

Anonymous said...

Damn, LG has punched Intel too hard. So far, Moorestown based smart phone product is only a manufacturer's prototype. While AMD is developing its Bobcat for several manufacturers with several SOCs configurations.

SPARKS said...

It seems a “Can full of whoop ass” has just kicked himself in the ass once again.

(Did I, or did I not, say they should lock that megalomaniac in his office?)

First, AMD/ATI has beaten the snot out of the marvel of engineering called ‘Fermi’ with the fabulous 5870 (to which I am a very happy owner) and the 5890.

Secondly, delays, yield issues, delays, ad nausea.

Third, heat, heat, heat, did I say heat? So much so they had to !UNDERCLOCK! (Filthy word) their HPC products to the tune of nearly 20%! Yuk.

According to Charlie’s report SuperMicro and NVDA are trying to head off a disaster.

Mr. ‘Who Flung Shit’ was going to storm the market and take on both AMD and INTC in HPC. A few months back he smiling like a Cheshire cat when INTC pulled back Larrabee.

Surprise
Surprise
Surprise

Notably, AMD was wise to stay the hell out of the fray and produce a terrific, reasonably priced product, well done.

Charlie’s in depth analysis says it all, especially the last paragraph.

SPARKS

http://www.semiaccurate.com/2010/05/05/
nvidia-downgrades-tesla-again/

Anonymous said...

The tick has struck and it looks decent. The tock is a coming and it will have 32nm on it, another doubling of density and even more power reduction and performance enhancement. ARM will still be stuck on 45nm and without highK metal Gate.

One more tick/tock and even Apple will have no choice but got where they went for their macbooks

Anonymous said...

It seems Mr. Tick Tock aka Monpoloy Man is very confident with Intel's manufacturing. Let's see if consumers want a smartphone based on performance alone. Globalfoundries is ramping its 32 nm SOI with full capacity before end of this year. At least OEMs have shown their rebellion to Intel although Intel can defend itself using some Clowns as its OEMs partners.

SPARKS said...

That’s it, I’ve had enough.

STFU, you flaming jackass, you don’t have a clue. You wanted a rise? I’ll indulge you, while others here simply (and far more wisely) ignore you.

From the GF website, they don’t even have 32nm on their roadmap, let alone 32nm on SOI.

http://globalfoundries.com/technology/
advanced_tech.aspx

Also, as far as SOI w/Hk on 32 is concerned, GF by their own admission, doesn't have a clue…



“GLOBALFOUNDRIES' ability to offer advanced process technology is powered by its highly successful and efficient collaborative R&D model centered on the joint development alliance (JDA) with seven world leading product companies including Freescale, IBM, Infineon, NEC, ST, Samsung, and Toshiba. With this alliance, GLOBALFOUNDRIES is driving the global standard for new technologies such as High K Metal Gate, with risk production at 32nm and 28nm in 2010. As a founding member of the JDA starting from 90nm, GLOBALFOUNDRIES has been able to leverage the collaborative R&D model to reduce the cost of process R&D that is projected to reach more than $1B US for 22nm and below.”

http://globalfoundries.com/technology/


How do you define “risk production at 32nm”? Ill tell you. The answer is here.


http://www.semiaccurate.com/2010/02/17/
nvidias-fermigtx480-broken-and-unfixable/


Sorry folks, I couldn’t help myself.


SPARKS

InTheKnow said...

Sorry folks, I couldn’t help myself.

Yeah, I was tempted to point out that no one has seen bobcat, but Moorestown has at least shown us functioning prototypes.

But since you asked a real question (inadvertently or not) I thought I should answer it.

How do you define “risk production at 32nm”?

Risk production is actually a term used in the industry to refer to the tail end of the start up process. Once first silicon has completed processing and met all electrical test parameters, factories usually begin ramping production. The fly in the ointment here is that the reliability analysis hasn't been completed yet.

For printed circuit boards, this takes something in the 6-8 week range. I expect it is similar for the semiconductors that go on the boards as well. You also need to get those first starts packaged, which probably takes as long as it did to make the chip in the first place. So there is the chance that a reliability issue will show up and you will scrap some or all of the material you produced up to that point. Hence the term "risk production" or "risk starts".

If GF has started making risk starts, that means they have a product that meets all their design targets for electrical parameters.

You can see from this explanation that the article on Fermi you linked to is a poor example of what can go wrong with risk starts. Those were problems that could be and were identified by electrical testing. Those issues are more of a DFM problem. A better example of the problems with risk production is something like Nvidia's cracking problem. Those parts looked fine until you ran them for a while, then they croaked. That is a problem that requires reliability testing to identify.

Anonymous said...

Yawn, lots of people are claiming to be in production, risk production, have Yields.

Look at silly TSMC, oppos chamber matching? WTF is chambermatching, how hard can that be? Tells you the problems go far deeper than simply targeting two deposition or etch chambers

Where the hell is that simpler and cheaper and easier to integrate gate first HighK process that was annouced more then two years ago. Today I will wager there is NO functional working CPU from that phantom annoucement.

In one corner you have shipped more then 1/2 billion chips on 45nm and probably 10's of million a month soon on 32nm.

Processing is getting harder and harder. It ain't so simple to just find some sugar daddy to pony up a few billion a year for 5 years and go on a shopping spree at ASML, AMAT. In the end you got to put it all togather. The recipe and secret sauce is getting harder and harder. Traditional oxide ain't going to cut it, and NO one got an answer except one company.. what does that mean?

to be continued

Anonymous said...

That the company that is ahead and making billions has more advantage then just the profit. It has process learning to feed forward for even faster learning on an even more complex situation at 22nm. Its a beautiful cycle where those "in the know" learn more and faster and those that don't wander around aimless faster and faster

Sure some company can hire some clever designers and throw a simple and very portable design togather at 65nm and scale it to 45nm. But wait those manufacturing companies really haven't ever shown they can do breakthru stuff. They just go shopping for tools. The company they are hoping to show them the way has taken the wrong path now on 4 huge things; Low K backend, strain, SOI, and now High K. You really think this company knows how to do things. I don't know where it leaves those companies betting on ARM.

That Moorestown platform from that Moore's law company looks pretty good. One more turn of the clock on 32nm and it'll be same story as the server guys.

Tick Tock tick tock

SPARKS said...

“probably 10's of million a month soon on 32nm.”

No “probably” here, bank on it. INTC is reporting shortages on 32nm laptop processors. In fact, according to ARS TECHNICA, Big Paulie has admitted they are “slightly behind” due to unexpected demand.

Apparently, the power sipping processors are closing the gap between netbooks and notebooks, my guess anyway. Perhaps we’ll see an expedited ramp up on 32.

Hear that Orthogonal, your work is cut out for you.

SPARKS

http://arstechnica.com/business/news/2010/04/
shortage-of-new-intel-laptop-chips-may-mean-laptop-delays.ars

Orthogonal said...

"Hear that Orthogonal, your work is cut out for you."

You have no idea. ;)

Anonymous said...

Ortho why are you reading this ramp baby ramp baby ramp!

SPARKS said...

All work and no play makes Ortho a dull boy.

SPARKS

Anonymous said...

You have no idea. ;)

I knew that, it's better for Intel to ramp up volume production for native CPU+GPU like Sandy Bridge than manufacture current solution like Arrandale.

SPARKS said...

Why?


SPARKS

InTheKnow said...

I knew that, it's better for Intel to ramp up volume production for native CPU+GPU like Sandy Bridge than manufacture current solution like Arrandale.

Sparks your anonymous poster is clueless.

You don't ramp the leading edge design on a new factory during qualification if you have the luxury of owning more than one fab on a given node. You ramp a well established design that has a large baseline data set to compare to. That way you can catch any issues in the new fab as quickly as possible.

Fab32 (I believe that is where Ortho said he works) may switch over to Sandy Bridge soon after start up, but they won't be qualifying on it.

SPARKS said...

Precisely, he stuck his foot in it, didn’t he? On one hand I’m reporting that there are shortages due to high demand, Big Paulie (and Ortho) are verifying it. By that knuckle heads estimation INTC should simply ignore a hot selling, high demand product.

Duh.

BTW, I did a search on the ‘Bobcat’ you mentioned on a previous post. They’re “sampling” a whole lot, but that’s about it. No working silicon reported, yet.
Well, they have six months to pull something out of the hat. The question is who’s going to do the process, GF, TSMC or UMC.

From what I’m reading it will be on bulk, as “G” pointed out in previous post.

http://laptopreviewshop.com/amd-ontario-system-on-chip-built-on-40nm-lithography-process.html

SPARKS

InTheKnow said...

For those interested in what Moorestown brings to the table, I refer you to this link.

I was pleased to see that Intel was upfront about where they are weak as well as where they have advantages. Overall, I think that Intel makes a pretty good case for Atom in the smartphone space.

Anonymous said...

Atom + one generation lead on process gets the platform very close, with one more spin of the Atom on 32nm and I think they will match ARM on simple metrics but have all the advantages of x86 that won them the server and all of the PC space but.......

Will the ship have sailed and in two years ARM be good enough. Lets be serious in a phone there is only so much performance one needs for a 3" touch screen. Sorry HD graphics, shooter games, excel/powerpoint editing blah blah blah don't make a compelling reason to convert if ARM is going to be good enough.

Did Intel make the wrong bet with Atom 4 years ago and not push to the lowest power. Should INTEL have never sold their x-scale... hmmm decisions that will haunt Craig Barret forever

InTheKnow said...

Will the ship have sailed and in two years ARM be good enough. Lets be serious in a phone there is only so much performance one needs for a 3" touch screen. Sorry HD graphics, shooter games, excel/powerpoint editing blah blah blah don't make a compelling reason to convert if ARM is going to be good enough.

That is the concern that I've been voicing for a while. I'm not convinced that Intel is moving fast enough. However, I think that Atom does offer a couple of key advantages that you missed.

First, ARM's performance is only close to Atom if you look as single tasks. Many ARM implementations will only allow you to run a single task. If you can run multiple tasks the ARM systems really bog down. I agree that you won't be looking at more than one app at a time on a smartphone. But the ability to switch quickly between apps is a big deal. And right now that is a big advantage for Atom.

I'm sure that ARM is looking to address this. What you need to remember is that in order for ARM to increase performance they will have to increase power consumption. It is basic physics and there is no free lunch.

The question here is whether or not Intel can bring down the power requirement faster than ARM can figure out how to bring up performance without moving out of the power envelope? This is where I'm not sure Intel is moving quickly enough.

Second, and I think more importantly, Atom has an advantage that stems from the existing computer architecture. For this statement to make sense you have to understand my viewpoint. I envision a user that has a mobile device they take with them. They synch it up with a more powerful home system and move apps freely between the mobile system and the home system. Apps will need to be able to transfer information seamlessly and quickly for this to be viable.

I understand that the apps will most likely have a different look and feel on the mobile device and the home system. But I believe the underlying architecture will be key to portability. From a developers point of view, it is going to be easier to develop apps for a single architecture than for multiple architectures. I don't envision ARM being able to power the home system in this scenario anytime soon. So Atom has an inherent advantage in being able to provide a single unified architecture with this home system running x86 architecture.

I ultimately believe that it is the unified architecture that will result in Atom being successful. But by being late to the game the fight will be longer, harder and more costly.

InTheKnow said...

Did Intel make the wrong bet with Atom 4 years ago and not push to the lowest power. Should INTEL have never sold their x-scale... hmmm decisions that will haunt Craig Barret forever

No, ARM is not a viable business model for Intel. There is no product differentiation for Intel going that route and they give up too much of their margins. Win or lose, Atom was the right choice for Intel.

SPARKS said...

ITK-Big Paulie makes a very compelling arguement regarding ARM and other architectures.

http://www.theinquirer.net/inquirer
/news/1648485/intel-silicon-future

SPARKS said...

Ah----OK----personally, I think this is where the real money is, full function, ultra thin, long life, etc.

http://www.anandtech.com/show/3737/intel-ultra-thin-culv-arrandale-boosts-performance-and-reduces-size

SPARKS

InTheKnow said...

Ah----OK----personally, I think this is where the real money is, full function, ultra thin, long life, etc.

I think you are right for the short term. However, if you buy into the idea of convergence that Intel is selling, you will see that this is not a long term solution.

As I interpret Intel's vision, ultimately, the small form factor portable device will go with you wherever you go. It will have computing power equal to or better than today's laptops. As you return home you will use it to turn on appliances, lights, etc. ahead of your arrival. Once home it will interface with your TV, and other CE devices and synch with your home system.

When you leave again you will have pulled all you need for the coming day from the home system and may even maintain a wireless link to the home system. In short, the portable device is the key that ties everything else together.

Will it actually play out that way? Who knows? But I do believe that is the image of the future that Intel is driving towards. If it does play out that way, whoever owns the portable device owns everything else. So while it may not be the most expensive piece of the puzzle, it will be what controls where the real money lies.

InTheKnow said...

Looking at this demo, it looks like Intel may have gotten it's integrated graphics to the point where they might actually be "good enough" for most applications.

I do wish they had picked something with a little more action though. This kind of raises the specter of the Larrabee demo, with not much going on. And we all know how that played out.

SPARKS said...

Can a group throw enough money at FAB tech to become a major world player in the long term?

Obviously, we shall see.

http://www.semiaccurate.com/
2010/06/01/global-foundaries-
expands/

SPARKS

J said...

@Sparks, that's not exactly new, just a refinement of previous generations.

SPARKS said...

ITK, this board's for you.

From what I'm reading DFI is shifting everything towards these embedded platforms. Hmmm, robotics anyone?

SPARKS

http://download.intel.com/pressroom
/kits/events/idfspr_2010
/pdfs/Tech_Insight-Tunnel_Creek.pdf

Anonymous said...

The clock is ticking for ARM

Wasn't it only a few short years ago that some big companies got togather forming a can't lose consortium to take on the elephant that can't dance.

Why is another consortium being formed to do battle with the same elephant?

Consortiums are formed when companies know they are in trouble, don't have enough money or know that they can't win alone. They are scared very scared. Why do companies and executives with big egos compromise, because they are scared very scared.

One more Tick Tock and ARM is finished. All things that beging with i will have the elephants silicon inside.

InTheKnow said...

Microprocessor Reports offers the following assessment of Moorestown.

Moorestown is a significant step for Intel. As the first x86 chip set suitable for smartphones, it establishes a foothold in an important market. Microprocessor Report concludes, however, that Moorestown still trails its competitors in both power and integration.

The full article (if you have access) gives a lot more detail around what they base their conclusions on. Overall I felt the arguments were reasonable and balanced and I'm inclined to agree with their conclusion.

They go on to say that they believe that Medfield (due out next year) will be pretty much a dead heat with dual core A9 performance wise. For Intel to succeed here they will need to do more than offer an equivalent solution. They will have to give customers a compelling reason to change to a new architecture. Atom's advantage (aside from the continuum argument) as I see it is going to be have to be performance.

The one clear advantage I see is that Atom seems to have the power to multitask while the ARM solutions don't seem to offer the same level of multitasking. The Microprocessor Report piece didn't evaluate that aspect, so they may be missing Intel's one big advantage.

On a related note, I heard a rumor that the LG phone was dropped, not due to hardware issues, but due to poor software support. I'm repeating the rumor here only because Intel's reputation for driver support is rather poor, so the rumor seems plausible to me.

If the rumor is true, Intel needs to get their act together on the software side of the house quickly, or this whole thing could still turn south on them. There is more to the ARM ecosystem than just a good processor, and Intel knows it. That is why they are licensing designs for many of the subsystems in Moorestown.

Anonymous said...

Like the clock is ticking, only one full cycle and INTEL is even by 15nm who will be in the game?

TSMC, Samsung, Global Floundring, and the constortium are so far behind on technology. They've compromised and followed IBM down the gate first path. What a mistake and two year detour. They are not just behind but really behind now. Time is learning, time is money, time is leadership and they are behind in all of that.

Sure ARM is everywhere and has the power leadership at the moment. Has every company on earth using it, does that sound like RISC a few decades ago.

But who is coming up in the reareview mirror? Who has the afterburners on? Who has superior process technology, superior manufacturing, billions in cash flow to fund new technology, new factories, new designs.

There are no shortcuts, he who has the most advance process with the lowest leakage wins. With another two design spins ARM will be behind. It will be hard for companies to stay on ARM just like it was impossible for them to stay on 6800, on PowerPC, on well you know.

Its impossible to resist, ARM knows it and that is why the consortiums were formed, they are formed when people know they are going to lose!

Anonymous said...

fiddle dee fiddle dum today Korea claims it got HighK metal gate too... AMD/GF and IBM we are waiting, didn't you do a vapor press release more than two years ago. Them Koreans have been hard at work and maybe past you right by!

It'll be interesting as it is gate first and we already know that to be first sometimes results in one being last if you know what I mean ;)

Anonymous said...

That's a low power (low performance) process, so it's not even really comparable to what Intel is doing. The high performance variant won't be "introduced" (whatever the hell that means) until ~end of Sept.

Samsung seems to have taken a cue from IBM and AMD and use nebulous terms like "introduce" instead of saying revenue product shipping (which would seem to be a better way to define a timeline)

Anonymous said...

http://www.electroiq.com/index/display/semiconductors-article-display/1174922828/articles/solid-state-technology/semiconductors/device-architecture/interconnects/2010/june/iitc-day_1__3d_tsv.html

http://download.intel.com/pressroom/archive/reference/ISC_2010_Skaugen_keynote.pdf I believe that during the 90nm/65nm days, Intel had high-k as 1 of the prominent "possible future solutions" to the next level. So based on that, from slide 8, we can expect computational lithography, 3-d, copper barrier films, iii-v, nanowires, optical interconnects, and carbon nanotubes as the next frontier.

InTheKnow said...

For those interested in where technology may be heading you can check out this paper and the accompanying slides.

There is a lot of highly technical detail in the slides, but it is pretty interesting if you are willing to dig enough to be able to follow it.

InTheKnow said...

Hmmm... For some reason the link keeps trying to direct to an htpps instead of http. If you are having trouble with the links try pasting this into your browser for the paper.

http://download.intel.com/pressroom/pdf/kkuhn/Kuhn_IWCE_invited_text.pdf

And this for the slides

http://download.intel.com/pressroom/pdf/kkuhn/Kuhn_IWCE_invited_slides.pdf

J said...

http://www.semiaccurate.com/forums/showthread.php?t=2731 I am reading this thread and people are claiming Bulldozer will be smoking, what do the smart people here think?

InTheKnow said...

Sorry, nomical, I don't have a clue whether Bulldozer will be any good or not.

First, I haven't seen any data on the 32nm process it is going to be manufactured on. Even the best design on a crummy process is going to be less than stellar.

Second, I'm a process guy. While I might be willing to evaluate Bulldozer's potential based on process data, I leave the architectural evaluation to others.

Anonymous said...

Intel q2 earnings:
$0.51/share vs 0.43 consensus
Net income: 2.89Bil
Revenue 10.77Bil vs 10.25Bil consensus
Margin for the quarter was 67%

Looking Forward:
- Projecting Q3 of 11.2-12Bil Revenue range (consensus midpoint was 10.9Bil)
- Increase yearly margin by 2 points (64-68% vd 62-66% previous guidance)
- Bumped up capex (fab capital spending) by an additional 400mil

Shares were up over 7% in after hours trading (needless to say both strong results and a strong forecast)

InTheKnow said...

I'm hearing rumors that GF's 32nm process is having teething pains and that AMD's Llano processor is suffering for it. Unfortunately, I haven't heard any details about what the specific issues are.

InTheKnow said...
This comment has been removed by the author.
SPARKS said...

Indeed, it is a fine time to be a long term INTC shareholder. Combine these results with a 3% EPS annually, INTC has performed well (and has gone the distance) in the face of this miserable economic pigsty.

SPARKS

SPARKS said...

ITK, typical AMD crap. Everything is rosy until the HV silicon comes out of the oven.

SPARKS

Anonymous said...

Record revenue, record profits, 66 points gross margin.

Damm if Bk looked so good let INTEL continue to march down the BK road. WTF is Sharikou and Scientia doing in their grandma's basement doing these days?

The power of scaling, the power of being first is really paying off. Invest them billions in new factories, new process and new designs leads to bigger profits bigger leads and even bigger profits.

This strategy crushed all other companies and architectures, damm even crushed their own itanic.

AMD is long gone, finished just like i predicted

ARM is next

Those that think it is bad think again. You need to profits to advance the process and give you more value.

Where is AMD / GF and the arabs.

Tick tock tick tock

InTheKnow said...

It will be interesting to see how this plays out.

GF is planning on being an early adopter of EUV litho. According to the statement in the article, they are going to be skipping the purchase of a pre-production tool. I suspect this is a nice bit of marketing spin as the IBM fab club will certainly have a pre-production tool to work with. So GF will get experience with a pre-production tool. They just won't be buying on for their NY fab.

SPARKS said...

It pains me to beat this one to death, but I think I’ll survive.

Regarding INTC’s 3% annual EPS, just today a commentary from another anal-ist. This one, at the Wall Street Journal, was weighing in on INTC’s future prospects. (Naturally, AFTER the $1.50 jump) He capped his commentary by calling INTC’s annual EPS “behemoth”.

Ya really got to hand to those WSJ anal-ists. They really know how and when to spot a winner.

And G, I was way off on the $25 by 2Q mark. I thought we’d be out of this financial cataclysm by now. I wasn’t smart enough to factor in EURO economics.

(This is why I use linesman pliers to make a living.)

SPARKS

SPARKS said...

ITK, from that link……….
“ he is hoping that EUV lithography will reduce manufacturing costs.”

Huh???? What are they going to make the lenses with rock salt?

SPARKS

Anonymous said...

A wise man says he who goes first to new tools pays extreme in price in yields, in productivity.

Has anyone purchased yet a Gate first device, anyone?

Anonymous said...

IBM/GF claimed a 32nm (or 22?) two layer structure done with EUV which now was clearly meaningless (much like the SiLK claims when they were originally made).

They may want to do production in 2014/2015, but there is not a clear path yet on how to get there which is why the tech keeps slipping one node at a time. You can't always schedule breakthroughs on the calendar, and this technology is not yet at a stage where it is merely engineering and debugging (which you can put a timeline on)

It is similar to high K... they had some short of short loop or researchy type claim on high K and how it was an option for 45nm, but everyone and their brother in the industry knew the "45nm option" was an utter joke and was only mentioned so it did not look like they were that far behind Intel. Even Intel wanted to originally put it in by the 65nm node, but the breakthroughs took longer than expected.

Orthogonal said...

"Has anyone purchased yet a Gate first device, anyone?"

Billions have, I presume you meant High-K gate first ;)

Q2 numbers were great. Q3/Q4 to be even better. Paul was modest in the conference call (although analysts were bizarrely down-beat all things considered). Ramp is incredibly aggressive.

Anonymous said...

I'd expect AMD to perhaps go slightly back to a loss in Q2 and maybe be profitable again in H210. April 7

AMD results:
Net loss: 43mil, $0.06/share
Revenue: 1.65Bil up 40% from last year, up slightly from Q1
Margin:45% (down from last quarter)

The margin seems odd since revenue went up, which often will help margins... oh wait the Q1 # was cooked by 4% by including some one time charges and benefits and was really 43%.

Dirk Meyer: "Intel clearly knocked the ball out of the park in their server business ... clearly we lost some [market] share."

Anonymous said...

From the AMD analyst call:

- AMD equity share of GF is now down to 28% as they did not match capital calls... (I'd expect ownership to continue to drop as AMD is still eating some of the loss as a part owner).

- Shockingly AMD chose to compare this quarter's Margin to last quarter's non-GAAP margin of 43% to show an increase. As they have done repeatedly, they roll out the high # (47%) when giving the results to make them look good, then when they need to make a comparison to a past quarter they use the 'real' #... this is pretty slimy. (They should consistently use either the GAAP or non-GAAP #, not whichever one makes things look good at the time)

- Meanwhile, we have seen the rate of yield learning below our plants on 32-nanometer, and as a result of the opportunity, frankly that Ontario gives us and the reaction to the yield situation, we've put our engineering resources into creating samples for Ontario supporting ODMs and OEM system designs, while we take a bit more time to work the 32-nanometer yields up the curve. (translation - 32nm is delayed enough that they've re-allocated resources to work on the 40nm Ontario netbook product which is being done out of TSMC... that does not sound good for 32nm)

Anonymous said...

AMD yield ramp delayed..

Deju Vu, yawn, who cares

Many moon ago it mattered more but today AMD is a nothing

This proves that even with Arab money thrown left and right that hard working and likely very good engineers shackeled in a bad business model directed by bad technologists at a consortium with infinite money will still produce garbage.

I can't wait for the delays when they get to the real hard stuff.

AMD may not be going bk but with crappy yields and being late might as well be

the clock has run out on them and GF

InTheKnow said...

You can't always schedule breakthroughs on the calendar, and this technology is not yet at a stage where it is merely engineering and debugging (which you can put a timeline on)

That's why I thought the link was interesting. You have a VP in charge of R&D promising to deliver material manufactured with a specific technology 4-5 years out.

AMD have said "nobody buys nanometers" how many times? Would it be fair to say "nobody buys your litho tech" in response to this?

His comments really strikes me as someone trying to find something to impress people with. The problem is that the tool you use is really only important if it adds some sort of competitive advantage. I don't see how an early technology, problematic, process tool is going to add the competitive edge. Better litho is not equivalent to a process advance like HK/MG. Once the litho tool is on the market, anyone with the cash can buy one.

InTheKnow said...
This comment has been removed by the author.
Anonymous said...

Making a chip is like cooking!

Everyone can go out and buy the most expensive cooking equipment.

ASML, Nikon, TEL, Applied etc. will say any tool to any body. Sure their biggest customers get first cut at the prototype but in the end the edge Intel has is in the cooks in the kitchen.

With big buck Arabs I'm surprised how awful and late the consortium is. Says somethign doesn't it. You'd think the best minds from AMD, Samsung, TSMC, IBM, etc.. with all their pooled money in 3 years could catch up.

Maybe gate first HighK aint' so simply and elegant. Does this sound like SILK round two?

SPARKS said...

Apparently, Dirk has inherited (and mastered) his mentors (Wreaktor Ruinz) penchant for AMD speak. I very much missed those interpretative comments you handled so readily in the old days. At times it’s hard for us minions in the trenches to decipher what the hell those corporate boys are talking about.

Elsewhere, added to the list of successful American Companies on the Democrat/Obama hit list, Golden Sacks was given a token slap on the wrist to the tune of a half a billion (a mere 14 days profit) for its error of “improperly informing” investors.

Sacks more than made up for the loss with a 7 percent rise in share price. At 75B in market cap that ain’t choke change. GOD if I could only afford 150 bucks a share.

I suspect INTC will be given the same. (Share IP my ass. On that count, I hope the Big Paulie and the boys fight the bastards tooth and nail.)

SPARKS

Anonymous said...

Intel reportedly settling with FTC.

http://www.eetimes.com/electronics-news/4204693/Report--Intel-settles-with-FTC

Reported fine: $0

The deal will involve increased FTC regulation of volume discount.

[/begin rant]

Gotta love the US gov't where the answer to everything is incerease regulation.

Banking crisis - we need more regulation (please don't consider the fact that we weren't enforcing existing regulations which in part enabled the banks to leverage themselves to greater than 30:1). Why would you want to enforce regulations when the real estate market was growing out of control, creating jobs at an unsustainable rate and generating lots of tax revenue?

BP oil spill crisis... increased offshore regulations. Yeah the regulators may not actually be inspecting offshore rigs at the frequency they are supposed to be (details, details), but with more regulations that will fix the ability to execute/enforce regulations (why wouldn't it?).

The problem: more regulations = more special interest on boith sides of a crisis = more political campaign money + more gov't jobs + more power + the ability to fix the next crisis. This means more money spent on 'fixing' the problem with a bandaid that is politically expedient rather than actually committing the resources to fundamentally addressing the issue.

The fundamental issue is the inability to actually enforce existing regulations (whether it's competence, money, political expediency) and that means the likelihood of more complex regulation succeeding is minimal. It does employ the political machinery and feed the campaign machine.

Why not put a 10th roof on that house with a leak rather than asking how do we actually fix it? Why spend the time to actually understand and address the issue when you can create new roofing jobs, have people lobby over how many roofs can be safely stacked on and how we should inspect it. And then when that roof leaks again in a year or two you can ask do we need new regulations and can we put on an 11th roof without the house collapsing?

[/end rant]

SPARKS said...

“ 'fixing' the problem with a "bandaid that is politically expedient rather than actually committing the resources to fundamentally addressing the issue."

This well defined, magnificently put, definition of common sense and logic is way beyond anything Washington (or Albany for that matter) could ever understand conceptually.

Smaller government and all the subtle nuances defined, my dear old friend, lays in peace with my beloved President Ronald Reagan.

God rest his soul.

It’s too late, too big, and those days are over.

It’s all simply out of control.

SPARKS

Tonus said...

"Why not put a 10th roof on that house with a leak rather than asking how do we actually fix it?"

As you noted, what the government actually does is promise to build a 10th roof and allocated the resources for it, then doesn't bother. There is an agency that is tasked with regulating offshore drilling, but it wasn't doing its job and apparently was being effectively run by the oil industry. Obama promised that he would clean up that particular mess, but he gets his share of lobbyist dollars from the oil industry and so nothing was done.

Politicians promise to add regulation only as a way to satisfy angry constituents and spend more tax money. Actually regulating is not really part of the equation for them. Just another reason why we spend so much, run up so much debt, and seem to get so little in return.

Anonymous said...

I'm not against regulation, it's necessary and if done right is very important.

The key though is when you put in new regulations, you have to clean up the old stuff at the same time instead of just leaving legacy stuff behind because there is no glory/value/visbility in removing old stuff. I look at regulations ("standards") in the semiconductor industry, and while there is some legacy stuff, generally speaking new safety, automation, and other stamdards/regulations get cleaned up as new stuff comes out.

Part of the issue in Washington is that it is no longer a representative government and is primarily filled by lawyers and career politicians who have never worked a day in their life out of the political arena. How many former engineers, scientists, truck drivers, farmers, etc are in Congress? Lawyers know how to interpret law (which is a necessary and important skill), but they are not problem solvers who know enough to understand oil rig regulations - and because of this they are reliant on special interest and lobbyists (or worse - their national party) to tell them what is necessary in most areas. You think even 2% of Congress has a clue how 'national' health care works or how to reform it?

This is why the FTC is essentially useless - they will create new hoops for Intel to jump through, they will not look at any of the old hoops and there will be enough confusion from them trying to enforce every hoop that they will be effectively useless (which is either a good or bad thing depending on how you feel about Intel's business practices)

At some point there needs to be a bridge period where work is started on a new, ground up set of regulations, and the old stuff phase out completely (as opposed to patched). Whether you're talking banking, tax code, real estate - everyone of these has become what the commercial industry became prior to failure.... too big to fail, too big to change or start over.

Could you imagine a tax system where you just paid taxes on income (or revenue for businesses) earned? Of course you couldn't pay off special interest via tax code exemptions with that or try to influence the housing market so the economy grows unnaturally so your party had a better chance of getting re-elected.

Could you imagine a system where money being put into a national pension fund (also known as social security) was only used for pensions instead of just being ported over to the generic Washington slush fund and spent on whatever they like? What would happen if a CEO used a company's pension fund to pay for general operating expenses? How quickly would he be fired and/or be in jail? So who's accountable for raiding the national pension fund (social security), year in and year out?

Sorry for the tangent - I guess my point is that 'more regulation' for Intel means little or no actual change. While a fine also would have had no impact (unless truly massive) at least we are not porting money over from a fairly efficient business to the most inefficient business entity in the world (the US Gov't).

Khorgano said...

We don't need to get into a big political argument about regulations again, but suffice it to say. Regulations by any government agency are illogical and never necessary. There will only be two outcomes, neither good for the consumer. Industry insider's will lobby/control the regulatory body to grant special priviledges to themselves and place barriers for their competition. Or the regulators will disrupt the competitive process and destroy the supply/demand and profit/loss mechanism that is critical to keeping a market successful.

Further, no government fund or system can ever be successful since they lack all profit/loss mechanisms (known as economic calculation) since revenue is disassociated from expenses. They are illogical and will never work. No matter how altruistic or skilled a governing body may be, they are doomed to failure.

Anonymous said...

Can someone give me the cliff's note of what Scientia, Sharikou and the AMD masturbators are fantasizing to these days?

Tick tock tick tock the clock has run out on them

SPARKS said...

Would anyone a little smarter than me like to speculate what the FTC is up to?

Apparently they need “more time to consider a proposed settlement".

The way I see it INTC is playing hardball with the IP nonsense and the FTC is looking for another option. Obviously, they don’t want to let INTC off the hook completely unscathed.


Am I wrong here?

http://www.marketwatch.com/story/ftc-extends-temporary-withdrawl-of-case-vs-intel-2010-07-21?reflink=MW_news_stmp

SPARKS

SPARKS said...

Well we can add yet another successful American company to the Democrat/Obama hit list. Today DELL was slapped with a 100 million dollar fine courteously of the SEC. I wonder who’s next GOOGLE, perhaps?

With regard to the US’s penchant for strict regulatory sanctions as of late, I say we just hand over the entire industry to the Chinese or to the Koreans. Hell, is anything they do regulated by anyone? The Chinese can even gain access to the US DoD with nary a complaint from our illustrious Secretary of State.

Watch the computer industry go by way of the TV, Auto, and the High-fi industry.

You can't make this crap up.

http://channel.hexus.net/content/item.php?item=25684

SPARKS

InTheKnow said...

Looks like not everyone is ruling against Intel in their anti-trust suits. In the class-action law suit alleging harm to consumers (incidentally, the first time the allegations have been tested in a court of law) the first round has gone to Intel.

Anonymous said...

^ Yeah I saw that... it's one thing for AMD to make a claim that the volume discounts were harming AMD, but I'm not sure how the class action argument goes?

We were harmed because we paid less for a computer because Intel gave volume discounts to win more/maintain business..... yeah, we may have paid less, but....

Had Intel not done these discounts, we would have paid more and AMD would have been healthier so in 5-10 years when I buy another computer it may be cheaper? Which should just about offset the discount I got on this computer?

Are people really claiming they were harmed by Intel discounts to OEM's? I can understand the argument about harm to AMD(not that I necessarily agree with it), but I don't see the financial harm to someone buying a computer in that time period.

A Nonny Moose said...

Can someone give me the cliff's note of what Scientia, Sharikou are fantasizing to these days?

Last I heard, Scientia stopped updating his joke of a blog and his joke of a mod, in favor of playing WoW. Yep, you heard it correctly - ol' Sci has turned into a World of Warcrack addict :P.

Sharikook seems to have stopped his slavering, hydrophobic-like demonization of all things Intel and now just predictably spins the AMD news to the upside. For example, his last post was all about how AMD is bringing Ontario forward and thus demonstrating their mastery of CPU-ness, and conveniently ignoring the reasons as stated by AMD - that Llano is delayed due to GF's gate-first 32nm sucking billowing bags of wind.

SPARKS said...

Even with INTC’s stunning result for last quarter, JP Morgan decided to downgrade the company. What’s interesting to note is that NVDA (among others) was thrown into the mix. The semis took a serious hit today.

More significantly however, FBR downgraded both AMD and NVDA. NVDA’s target price is at 8 bucks a share. You don’t think TMSC “chamber matching” issues have anything to do with this downgrade, hmmm?
http://www.marketwatch.com/story/amd-nvidia-downgraded-to-market-perform-at-fbr-2010-07-29

NVDA’s power sucking miniature space heaters are not selling well either. AMD creamed them with the 5xxx series graphics cards over six months ago. NVDA has issued an 18 percent drop in revenue. I believe high end graphic cards have hit the wall, so to speak.

http://online.wsj.com/article/BT-CO-20100729-717965.html

In any event, I think those in the know (pun intended), plus our inside boys, are keenly aware that 32nM for other companies isn’t going to be as easy as INTC made it look.

The next five months should be very interesting with September (going back to school) and with December (Christmas) just around the corner.

AMD at 32, nah.

SPARKS

Anonymous said...

> suffice it to say. Regulations
> by any government agency are
> illogical and never necessary.

While I am a firm believer in the profit/loss principle, this is only half the story of what makes a market successful. I think this has been amply proven by financial events of the recent past. In fact, counting on only the profit/loss principle can have a destructive influence on the market in the long run.

> the regulators will disrupt the > competitive process and destroy > the supply/demand and
> profit/loss mechanism that is
> critical to keeping a market
> successful.
again you are demonstrating a biased view of what makes the market work

> Further, no government fund or
> system can ever be successful
> since they lack all profit/loss
> mechanisms (known as economic
> calculation) since revenue is
> disassociated from expenses.
> They are illogical and will
> never work.
huh, any empirical evidence to back this absolute statement?

SPARKS said...

“While I am a firm believer in the profit/loss principle, this is only half the story of what makes a market successful.”

Anon, your point was hard to target. While your passionate argument was genuine, its focus however, frankly, was nebulous at best. What precisely are you trying to say?

The key word here is “expound”. How do you think it should work?

Bust a move, brother.

SPARKS

Anonymous said...

"Anon, your point was hard to target. While your passionate argument was genuine, its focus however, frankly, was nebulous at best. What precisely are you trying to say?"

sorry but what did you find nebulous?

Khorgano said...

"While I am a firm believer in the profit/loss principle, this is only half the story of what makes a market successful. I think this has been amply proven by financial events of the recent past. In fact, counting on only the profit/loss principle can have a destructive influence on the market in the long run." Anon

A comment like this tells me you want to believe in free-markets but don't have all the pieces to understand how it works.

First, the markets were never un-regulated or de-regulated to begin with. There are numerous regulatory agencies from the FED, to the FTC, SEC, DoJ, US Treasury, Senate/Congressional banking and economic oversight committee's, and decades of laws and congressional Acts (i.e. Community reinvestment act) .

The latest financial collapse was by no means a "market failure", but clearly a failure of government oversight and intervention. People will blame it on unrestrained Capitalism but that's a load of bull crap. Here's why, you might say "The blame lies with speculators or sub-prime borrowers with bad credit or unscrupulous lenders or greedy bankers packaging loans into CDO's or the asleep at the wheel ratings agencies" yada yada yada. Did they play a part in the financial collapse? You betcha, but they were only proximate causes of the recession.

The ultimate cause is the Federal Reserve, more specifically, the regulation of interest rates caused the collapse, the regulation of which is the most economically destructive thing that can be done. By keeping interest rates artificially low, the Fed fueled all the "irrational exuberance" that led to the asset bubble and subsequent collapse.

In a TRUE free market, with NO regulation, the interest rates would be set by the market, not by a suit at the Fed. If ALL of the afformentioned "causes" of unregulated markets were to align and begin a new asset bubble like previously seen. The only way to get the Credit required to fuel the purchase of so many new/used houses would be to borrow the money on the open market from other people/firms. The HUGE demand for credit would make interest rates necessarily skyrocket and the sub-prime mortgages quickly become unprofitable, and soon after prime mortgages.

Free-Markets regulate themselves through interest rates and prices. Government regulation and intervention disrupt this process and cause the very things they are trying to stop.

"huh, any empirical evidence to back this absolute statement?"

Yeah, it's called Social Security and Medicare, both are insolvent, the US taxpayer is screwed.

SPARKS said...

“sorry but what did you find nebulous?”

You.

Your position, your point, your thrust, It’s very easy to criticize what someone says on a line by line basis, yet ironically you still haven’t backed YOUR criticism/comment with supporting “empirical evidence”.

Then you ask what’s not understood?

First, reread Korgano’s well constructed comment. (Hint: you can tell by his content he’s a bright guy who knows what he’s talking about)

Second, reread Korgano’s comment, and then read your original post.

GET IT?

SPARKS

Anonymous said...

"First, reread Korgano’s well constructed comment. (Hint: you can tell by his content he’s a bright guy who knows what he’s talking about)"

That is simply nonsense. Just because he has a certain vocabulary doesn't mean it is the gospel truth. In fact he makes many absolute statements without backing them up. And I am asking him to back up his statements.

What is nebulous about that?

InTheKnow said...

In a TRUE free market, with NO regulation, the interest rates would be set by the market, not by a suit at the Fed.

Khorgano, I usually don't have much to say on the financial discussions on this board, but in this case I'll make an exception. I believe your basic premise is flawed.


In your comment you seem to tacitly accept the existence of the Fed. But the Federal Reserve is not an entity created by free markets. It is an instrument designed solely for regulation of the market.

To ensure a completely free and unfettered market, you would have to eliminate the Federal Reserve and all other entities like it worldwide. In its place, you would have to implement some sort of species standard, be it gold, silver, energy, or any other tangible/measurable standard.

And I would postulate that a completely free and unfettered market would not necessarily be nirvana either. It would be far more dynamic than our current system resulting in local pockets of high unemployment (in a global sense) and potentially increase the gap between those that have and those that don't.

I'm not a socialist by any means, but let's not emulate the environmental nut jobs of the world here. All too often they focus on dealing with one specific issue without looking at the overall cost versus the benefits of a proposed change.

InTheKnow said...
This comment has been removed by the author.
Khorgano said...

ITK, I guess I didn't make myself clear. My personal belief is to see the complete abolishment of the FED. I do not "tacitly accept" its existence or believe that it is doing anybody any good (except for bankers/politicians and the well connected.)

We would then return to a hard asset currency (like gold/silver something else perhaps). There's no need to eliminate paper money, only FIAT money. Paper money would still be used since it is backed by a hard asset. The reason people used Gold historically is NOT because it has intrinsic value (intrinsic value does not exist), but that it has all the desirable characteristics of a "good" currency. Gold is not easily counterfitable (most important), durable, highly divisible and has a good relative scarcity.

"And I would postulate that a completely free and unfettered market would not necessarily be nirvana either."

Correct. No one said it would be. The point isn't that Free-Markets are Utopian or Nirvana. The point is that they eliminate monopolies and regulate greed. Greed is a part of the human condition, it can not be controlled by a government (Govt is a monopoly and attracts the kind of people you don't want having that kind of power). In a free-market, a "greedy" person can only make money by serving the market. Your bottom-line will be directly proportional to how well you serve customer's (since you don't have regulations barring your competitors giving you preferential treatment).

Free markets channel greed into productive forces. Governments and regulators channel greed for their own self interest.


It would be far more dynamic than our current system resulting in local pockets of high unemployment (in a global sense) and potentially increase the gap between those that have and those that don't.

Baseless assertion. However, we do have this now with State controlled economies.

Khorgano said...

Edit for clarification, should have read:

Gold is not easily counterfeitable (most important). It is durable, highly divisible and has a good relative scarcity.

InTheKnow said...

The point is that they eliminate monopolies....

I wholeheartedly disagree with this statement.

CPU manufacture is a prime example. The cost of entry is so high at this stage of the game that the main competitor (Intel), left unrestrained, is capable of completely cornering the market. AMD doesn't have the market cap to survive as long as Intel could squeeze them. Without anti-trust legislation, there is no reason on earth to believe that Intel wouldn't take the necessary steps to capture the remaining ~20% of the market they don't own.

I believe the logical conclusion from this example is the market creates monopolies in cases where the barriers to entry are high. Whether this is a good thing or a bad thing is another question, but the evidence would seem to contraindicate your statement.

InTheKnow said...
This comment has been removed by the author.
InTheKnow said...

Khorgano, since your economic theories seem to be based largely on the free market principles promoted by Adam Smith, I'll direct you to a quote from "The Wealth of Nations".

The trade of a joint stock company is always managed by a court of directors...The directors of such companies, however, being the managers rather of other people's money than of their own, it cannot well be expected that they should watch over it with the same anxious vigilance with which the partners in a private copartnery frequently watch over their own. Like the stewards of a rich man, they are apt to consider attention to small matters as not for their master's honour, and very easily give themselves a dispensation from having it. Negligence and profusion, therefore, must always prevail, more or less, in the management of the affairs of such a company.

A careful reading of Smith's work should also lead you to despise corporations as well as government intervention. All his economic theories (and, therefore, the derivatives based on his work) indicate that for his system to be realized, markets must be comprised of small groups of buyers and sellers. For good or ill, this is not the world we live in today. This difference must either be accounted for, or eliminated if we are to realize the state of affairs envisioned by Smith. Implementing only part of the solution invites disaster rather than improvement.

Khorgano said...

Capital is not a barrier to entry. Anybody can acquire capital on the market. There are entire industries devoted to providing capital to businesses. Yes, the amount of Capital is high to get into semi-conductors, but that's only because the amount already invested is enormous.

Lack of sufficient capital is not a barrier to entry, that can be overcome easily enough. However, A government regulator placing restrictions on where you can build, what you can make, who you have to pay off and how much. That is a barrier that is completely artificial and many times cannot be overcome.

Don't confuse a single source supplier (Intel if AMD were to BK) vs a Monopoly (Your electric utility company). A monopoly is a person, or group of persons that maintains absolute control over an industry/geography etc by force. A single source supplier is a market chosen entity based on superior products/efficiency/consumer preference etc. Intel would be the latter.

We've had this discussion before. Even if they do become one, prices will not skyrocket, they are very much subject to supply and demand. If prices go to high, that reduces the barrier to entry for competitors. Not to mention Intel is susceptable to the upgrade cycle. They can't sit on a technology, they have to innovate and reduce prices to survive.

Khorgano said...

Adam Smith was a classical economist but is not a free marketer, at least, not according to any free market theory promoted in the last 150 years. He blazed a lot of new territory, but was fundamentally flawed as demonstrated by Menger, Mises, Hayek and Rothbard.

"A careful reading of Smith's work should also lead you to despise corporations as well as government intervention. All his economic theories (and, therefore, the derivatives based on his work) indicate that for his system to be realized, markets must be comprised of small groups of buyers and sellers. For good or ill, this is not the world we live in today. This difference must either be accounted for, or eliminated if we are to realize the state of affairs envisioned by Smith. Implementing only part of the solution invites disaster rather than improvement."

What I think you meant by your comment is that State Capitalism should be despised, and I agree. The purpose of a free market is to eliminate priviledged status from groups.

We colloquially refer to business's as corporations and vice versa, but there is an important distinction between the two that make all the difference in a Free Market.

wow, I didn't think I'd get into this much theory here, but we have good discussions and questions.

A business is an abstract concept describing the people, capital goods and property working in concert for specific economic goals. A Corporation is a State Sponsored entity that is essentially legal reification. AKA, corporate personhood. It turns an abstract concept into a legal person. A Corporation is nothing more than a legal meatshield for the executives that run a business. (A la, BP and the oil spill?)

In a true free market, executives would be held responsible for their actions, not fictitious entities. This is the motivation to keep large firms in line and not cause harm to others.

As you said, implementing part of the solution leads to disaster, not improvement.

SPARKS said...

“Don't confuse a single source supplier (Intel if AMD were to BK) vs a Monopoly (Your electric utility company).”

“Lack of sufficient capital is not a barrier to entry, that can be overcome easily enough. However, A government regulator placing restrictions on where you can build, what you can make, who you have to pay off and how much. That is a barrier that is completely artificial and many times cannot be overcome”.

Oh, how true. This is not economic theory! Its economic facts damn it! I have living, solid “Empirical Evidence”, all of it, for over thirty years!

Single source supplier and monopoly!?! I’ve got two!
The Long Island Railroad, they are, as of late, asking for another rate increase to the tune of 10%. Always on the verge of BK, AND heavily subsided by both the State and the Federal Government.

The second is the Long Island Power Authority (Formally The Long Island Lighting Co.) Need I say more with boring details?

Oh, there is a third, or at least the New York legislators thought there might be, AMD.

LIPA
LIRR
AMD

These are prime examples of how government intervention can cost taxpayers billions of dollars and still have a myriad of functional/economic issues and problems.

They can’t throw enough money at these companies to keep them solvent.

Theory my ass, I’m living (and still paying) proof.

(Hey new Anon, how’s that for fancy vocabulary, it’s not about words, it’s about content, and I don’t have HALF the brains of some of these guys.)

SPARKS

SPARKS said...

“The settlement order, which takes effect immediately, also ensures that manufacturers of discrete graphics chips will be allowed "access" to Intel CPUs for six years. That means that Intel will be required to maintain an open, industry-standard PCI Express bus to connect to its CPUs—a significant change in the CPU and chipset landscape.”

Does this mean INTC has been forced to license chipsets design to NVDA? If so, who determines whether the price is right? And what the hell is “access”?

http://www.extremetech.com/
article2/0,2845,2367383,00.asp

SPARKS

Khorgano said...

Yes, it appears Intel is now banned from "Anti-competitive practices", which is government code words for, banning competitive practices in order to save competition.

Kind of like how Bush abandoned free-market principles in order to save the free-market.

Silly regulators, when will they ever learn.

Anonymous said...

"(Hey new Anon, how’s that for fancy vocabulary, it’s not about words, it’s about content, and I don’t have HALF the brains of some of these guys.)"

@Sparks - you always are entertaining, however I doubt if the points you make are relevant to the larger discussion. Nice try though.

SPARKS said...

From your original post ……

“> Further, no government fund or
> system can ever be successful
> since they lack all profit/loss
> mechanisms (known as economic
> calculation) since revenue is
> disassociated from expenses.
> They are illogical and will
> never work.
huh, any empirical evidence to back this absolute statement?”

From your last……….

“I doubt if the points you make are relevant to the larger discussion.”


Irrelevant you say, LIRR, LILCO, and AMD?

LILCO failed thanks to the governor of New York and heavy government “guidance’.
AMD gave up its foundries, gave up its IP, sold the lot in the backyard, and still has a crushing long term debt, even with subsides from state and foreign governments. AMD is government poster child for an orphaned failed company. The LIRR CONSTANTLY operates in the red and is held together with 60% federal and state money

FACT- IF each of these weren’t subsidized by government money and regulations “they would never work”. My commute, my utility bills, and my taxes, would be considerably less, personally.

What part don’t you understand, flamebait? It is you that has the penchant for irrelevancy.

(I’m done, all)

SPARKS

Anonymous said...

After reading these discussions, I think we should change the code of conducts in the US Government. We need to let the losers die or to be owned by the winners. If Paul Otellini is the US President, our hopes are becoming a reality. I believe when Intel's chip has been implanted in everybody's brain, our perfect world will benefit us as the winner.

SPARKS said...

“I believe when Intel's chip has been implanted in everybody's brain, our perfect world will benefit us as the winner.”



Short sighted fool; it’s that kind of myopic thinking that leads us back to medieval times of fear and despair. It’s amazing to me that such narrow minded thinking still exist in this day and age.

Have you considered embedded devices that could possibly regulate pituitary output function enabling millions sufferers to eliminate a lifetime of related problematic conditions?

http://www.umm.edu/endocrin/pitgland.htm


How about implanting other embedded devices to regulate glucose regulation? Perhaps you need a lesson in unregulated insulin production?

https://health.google.com/health/ref/Diabetes

How about thin film technology and processors to eliminate blindness due to the tragic results of macular degeneration?

http://www.mayoclinic.com/health/macular-degeneration/DS00284

I recently put rest a ninety three year old woman who suffered from the last two codirions of the three above. After caring for her for the past twenty plus years I can tell you from personal experience I would have more than embraced the technologies you stupidly dismissed categorically

Please, do us all a favor. Don’t go away mad, just go away.

SPARKS

Anonymous said...

why bother argue with this fellow. There are people posting in the forum, pretending to be a company's supporter, but giving really stupid comment in such with the purpose of protrarying extreme fanatic behaviour rest of the company supporters. If this fellow praises company 'I', you have to suspect this fellow to be company 'A' fanboy and vice versa.

Anonymous said...

Yeah, Spaarks and other posters are whining all the time because Intel should not be punished with its success and the losers like AMD and VIA don't deserve to be alive because of their own failure.

SPARKS said...

Understood.

Anonymous said...

Today is a sad day. Intels future is no longer clear and secure!

The executives have blown billions on another misguided adventure.

This will go down in history as one of the worst decisions in corporate history

The most dangerous thing is the koolaid being drunk by the intel folks. Damm if they were to come to me and offer me a 60% premium on my company I'd take it!

WTF does a software security company help INTEL. wake up your core competence is silicon, design killer complex x86 chips.

INTEL is getting its clock cleaned by ARM in he mobile space. They blow 7 billion to buy a software company. You think Apple, Samsung or other company is going to by a x86 mobile cellphone because it has McAffe bundled with it? The executives have taken their eyes off the ball, or maybe they think they can't win and are taking a different angle. Either way a very sad day!

InTheKnow said...

INTEL is getting its clock cleaned by ARM in he mobile space.

Intel has yet to put a legitimate competitor in the mobile space. It is hard to "get your clock cleaned" when you haven't even taken your best shot yet. Intel has a lot to learn about SOC's and they know it. If you can say that they are getting their clock cleaned a year from now with 32nm Atom SOCs on the market, I might agree.

Until then, Intel doesn't have anything compelling to offer in the smart phone space. 45nm Atom SOCs are competitive, but you don't unseat the dominant player in a market by being competitive. You need to be better.

Regarding the comment about the McAffee purchase, I'm at a loss. I'm sure they want to build the software capability into the hardware, but don't see how buying a software company helps them with that.

A Nonny Moose said...

From the news stories Intel & McAfee were cooperating on some sort of hardware/software security solution, and my guess is that Intel thinks it got some sort of killer device to ensure online security for cellphones to desktops.

A Nonny Moose said...

From Infotech: Intel Loses, AMD Gains as Processor Shipments Rise in the Second Quarter.

While Intel has a much bigger share of the worldwide microprocessor market, its performance in the second quarter was not up to the mark. As per market research firm International Data Corp.’s latest report Worldwide PC Microprocessor 2Q10 Vendor Shares, Intel lost 0.3 percent of the worldwide PC microprocessor market share in the second quarter, while rival Advanced Micro Devices gained 0.2 percent. And the third competitor VIA Technologies earned 0.3 percent. This loss was based on an overall worldwide unit market share of 80.7 percent as against AMD’s 19.0 percent share.

By same token, in 2Q10 unit share, Intel lost some ground to AMD in mobile processor market. The IDC report indicates that Intel earned 86.1 percent share in the mobile PC processor segment with a loss of 1.7 percent. And AMD finished the quarter with 13.7 percent share, a gain of 1.6 percent. While VIA earned 0.2 percent.

However, performance in the PC server/workstation processor segment is encouraging for Intel. In this sector, processor giant dominated with 93.5 percent market share with a gain of 3.3 percent, while rival AMD earned 6.5 percent, representing a loss of 3.3 percent. Similarly, in the desktop PC processor segment, Intel earned 72.2 percent, reflecting a gain of 0.5 percent. Here, AMD earned 27.3 percent, showing a loss of 0.7 percent. VIA earned 0.5 percent.

Besides strong server chip business, the reason for Intel’s gain in the server microprocessor market is being attributed to slow product transition at AMD.

Overall, in comparison to the first quarter 2010, the worldwide PC microprocessor unit shipments and revenues in the second calendar (2Q10) rose 3.6 percent and 6.2 percent, respectively, according to the latest PC processor study by IDC.

In a statement, IDC’s Shane Rau, director of Semiconductors: Personal Computing research, said, "Such a sequential increase in PC processor shipments alone would have been enough to conclude that the first half was strong for the market." "However, a modest rise in revenues, too, points directly to a rise in average selling prices. System makers bought more and higher-priced PC processors in 2Q10 than in 1Q10. Digging a little deeper into the numbers shows that they bought more mobile processors and more server processors, while desktop processors remained flat," added Rau.

He added, "Intel got significant sales traction for its new 32 nm Xeon DP products, and 45 nm Xeon MPs.”

In short, the IDC study shows that mobile PC processor unit shipments rose 6.5 percent quarter over quarter, PC server processors rose 6.1 percent quarter over quarter and desktop PC processors declined 0.1 percent quarter over quarter.

Looking forward, commented Rau, the second half of the year will be seasonal given the early build for Intel's Sandy Bridge and AMD's Fusion architecture launches, but lower than the year-over-year growth seen in the first half of the year. “2011 remains a wildcard in terms of sustainable unit growth," concluded Rau.


Seems to agree with what I've seen on the web as AMD is really flogging their mobile fusion solution. And not surprising that AMD lost yet more high-profit server marketshare, a fact alluded to by Meyer during the last conference call.

A Nonny Moose said...

By the way, over on AMDZone the zombie fanbois are gabbling & shrieking over AMD's 36% loss of server marketshare, and of course attacking the poor poster who brought the fact up. As usual, they just don't wanna hear nuttin' bad about AMD :).

Anonymous said...

Please, take the time to read the whole thing before making such stupid comments.

The guy stated that AMD lost market share because of competitive reasons when the truth is far from that.
Magny-Cours is AMD's new platform for servers, and this one will take time to establish just the same way socket-1207 did.

SPARKS said...

Moose, I read that too. It comes as no surprise, however, as we (here) knew this would happen over a year ago when INTC began its server assault in earnest.

As far as the numbers go, the INQ has their own take.

They also mention AMD’s increase in laptop market share. This was bound to happen. The combination of AMD’s cheap low end dogs and INTC’s ATOM net books no doubt affected Laptop sales overall. We predicted there would be a fine line between high end net books and low end laptops. (So did INTC) It’s a case where size definitely matters. Big, heavy, and low stamina is definitely not cool these days.

Conversely, cheap is definitely in, especially in this rag tag ecomony. A place where AMD can flourish, they don’t make money but they have market share. It’s business as usual.

Finally, as a INTC shareholder, I’ll take the high end, high margin server market any day (over 91%), and then low ball them to death with ATOM’s.


By the way, I never read a “stupid” comment from you, ever.

http://www.theinquirer.net/inquirer
/news/1728672/intel-grabs-amds-
server-share

SPARKS

A Nonny Moose said...

Anon: Please, take the time to read the whole thing before making such stupid comments.

The guy stated that AMD lost market share because of competitive reasons when the truth is far from that.
Magny-Cours is AMD's new platform for servers, and this one will take time to establish just the same way socket-1207 did.


Possible, but doesn't explain how AMD managed to lose ~80% of their server marketshare in just 3 years. As the AMDZone poster mentioned, AMD had around 30% of the server market in 2007, and in Q2 2010 they dropped to a bit over 6%.

IIRC Magny Cours was available from AMD before either the Nehalem EP or Westmere Xeons, so it had a longer time to establish itself in the market than either of the latter two. And with Bulldozer server coming up in about a year from now (assuming no further 32nm delays by GF), MC is getting into the "let's wait on BD" area, so IMO it doesn't have a lot of time remaining to be established in the market.

A Nonny Moose said...

Oops, make that Nehalem EX in my comment above :).

Sparks, read the Inq article and although they lay the blame on the server vendors slow to transition, makes you wonder why those OEMs would take so long to transition. Maybe they had too much leftover Opteron inventory they wanted to move first, or they are cautious about MC's prospects despite the low price? Anyway, not good for AMD since Intel had a blowout Q2 and AMD managed to lose money yet again (although that was due to their GF stake).

As for BD, it will be interesting to see if it gets pushed back like Llano did, due to GF's problems with their SOI gate-first HKMG 32nm node not ramping like it was expected to. I wonder if GF is devoting as much time and resources to it as they are to the 28nm bulk Si half-node, which has lots of profitable customers who can pay cold hard cash. ATIC is not stupid, as shown by their previous unilateral renegotiating the spinoff deal with AMD, to AMD's detriment due to falling stock price.

SPARKS said...

“….OEMs would take so long to transition. Maybe they had too much leftover Opteron inventory they wanted to move first, or they are cautious about MC's prospects despite the low price?”

That’s interesting; no doubt a combination of both, plus others as well. INTC’s server assault comparing core count vs. power and efficiency, power scaling, etc., could also be a factor.

Further, AMD’s hype doesn’t cut it anymore, remember CRAY years back?

Most importantly, even with the i7 socket change, I believe IT worldwide was ready for a quantum leap in hardware (EX) as opposed to a rehash of a dated architecture, no matter how many cores AMD threw at them on same Opteron platform. Really, how many times can you beat the same old horse before you realize it’s the same old horse? Meanwhile, all the new tech is blowing by you.

Can Advanced Micro Designs go toe to toe with INTC with R+D development, architecture and process combined? Nah.

The number show it and it’s not gonna stop. AMD, as so many on this blog predicted before me, will be relegated to a niche consumer market. Low end laptops, budget boxes, white box market, etc.

They can’t even hope for a lawsuit. That card has been played, too.

And let us not forget that long term debt.

Basically, it’s over.

SPARKS

SPARKS said...

Well, it seems while Advanced Micro Design is masturbating the masses with heaps of ‘Bull Dozer ’, Sandy Bridge has got the Tech press drooling, and (gasp) pretty good graphics too!

Sorry for this but I’ve got to agree with ‘ole Lex. The clock is ticking and Sandy Bridge is right on time.

http://www.anandtech.com/show/3871/the-sandy-bridge-preview-three-wins-in-a-row/2

More to the point, INTC is still executing flawlessly. To all involved, again, well done!

SPARKS

A Nonny Moose said...

Hmm, I'll wait for the benches on Bulldozer, but yes SB looks pretty good, even on the ES that Anand tested. I'm thinking 20% IPC improvement over Nehalem once the shipping models with working turbo are tested, and the integrated GPU looks pretty impressive too despite beta drivers and probably no turbo. The GPU was the smaller 6 execution unit too, not the 12 EU.

I'm also thinking part of AMD's Q2 gain in mobile marketshare was due to OEMs waiting on SB before refreshing their notebook lineup. Also AMD has been flogging their "fusion" notebook packages.

So those leaked benches a month or two ago, over on XtremeSystems that showed SB to be underwhelming, now look pretty suspect :). No wonder Dr. Who was threatening to withhold any future ES samples to certain people there :P.

Anonymous said...

It's a waste of time to argue with you guys, even if the evidence is shown on your noses.

But why should I expect better from an intel fan-based blog.

And I thought sharikou was bad, but this blog is worst with all the fud and misinformation floating around.

Tonus said...

"It's a waste of time to argue with you guys, even if the evidence is shown on your noses."

Somehow, I knew this recent exchange would end this way, because they always do. Some AMD fanboy troll makes a few comments, gets a response he doesn't like and cannot counter, and so he 'leaves' in a huff. Keep telling yourself that Nonny's clear and detailed reply to your post was just head-in-the-sand posturing if you want, I'm sure you'll be back for another beating in due time.

Anonymous said...

Bulldozer will be another shell game of AMD changing 2 variables at the same time (tech node and architecture) and assigning the gain to whichever variable suits them.

Even if BD ends up no better than K10 (I'm not suggesting this - it should be better), it should see 30% gain from process tech node transition alone.

My guess is it may give 10-20% gain and when combined with the process tech AMD will hail it as a breakthrough architecture and the press will ear it up as they won't be able to assess how much of it came from the process node. The question will be what would have been the case if they just migrated K10 to 32nm (to assess architecture vs process).

This is smart from a PR perspective as even if either the architecture or process transition underdelivers, by combining the 2 together there's a better chance of getting a good gain and not exposing an issue (if there is one). Of course from a technology and risk perspective a new architecture and new process is not exactly the best path to success - if there is a problem on say speed/power/binsplits it will make it that much harder to troubleshoot.

This is going to sound crazy, but anything less than a 45-50% improvement is going to be disappointing (30% process, 15% architecture)

SPARKS said...

Regarding PR posturing at AMD, they have decided to drop the ATI brand name. The reason you may ask? According to them, “Time and again, our research confirms that the average PC buyer is unaware of what processors are under the hood of their PC.”

Here’s the catch. They will keep the Radeon, Firepro, etc., monikers just in case, for people who know what’s under the hood.

A couple of things occurred to me. What the hell is an average PC buyer? Has ATI and their 5xxx recent successes outshined the mother company at AMD? (Surely, at this juncture AMD would like nothing more than to have brand recognition to go away in its entirety, especially on the processor side of things.)

Do they believe that the days of the high power graphics cards are coming to a close as they integrate fusion in the future?

Will powerful AMD cards give the ailing processor company more prestige and more clout?

Will future INTEL laptops sport the AMD logos and stickers as they sport ATI logos now? (Yeah, when hell freezes over.)

So, do the graphics card wars now turn from ATI vs. Nvidia to AMD vs. Nvidia?

Personally, as an enthusiast, I’m going to miss the “Red Team” as it turns to another pale shade of green. I have 2 5870’s in CrossFire X, they are both red, but I didn’t care who made them, or who got the money. I have purchased ATI exclusively for nearly 2 decades, I liked the power and price of the cards, even if they didn’t match NVDA’s best.

However at 400 bucks a piece I can definitely tell you two things. I know what’s” under my hood” and losing the ATI brand name is a mistake.

They spent 5B to kill a great company

SPARKS

http://blogs.amd.com/play/2010/08/30/evolving-the-amd-brand-portfolio/

SPARKS said...
This comment has been removed by the author.
SPARKS said...
This comment has been removed by the author.
SPARKS said...
This comment has been removed by the author.
SPARKS said...
This comment has been removed by the author.
Lem said...

Sparks, do you really think AMD killed ATI? It seems ATI's cards have been ever improving since AMD bought them. The HD6000 series appears likely to continue this trend.

SPARKS said...

Do I think AMD killed ATI? No, but they are killing the name.

Do they make great cards? Yes, as I said I own 2 and am ranked 8th on Sandra Benchmark. Better than NVDA in fact.

If you read the link you’ll see that this is all about killing the name ATI, and by my estimation, giving AMD more brand exposure with a superior product. They certainly haven’t had that under the AMD name for quite some time now, have they, despite what the poster on the link says.

Hey, perhaps the same people who don’t know what’s “under the hood” don’t know that ATI is AMD. Maybe the people who know what’s “under the hood” don’t know either.

They are going to get an education when AMD kills the ATI name.

Do you get my point?

SPARKS

Lem said...

Yeah I see your point Sparks, I was just questioning your last line, "They spent 5B to kill a great company". AMD obviously thinks they don't need the ATI name, which I don't think they do. For anyone who knows or cares, it'll be obvious who nVidia's main competitor is. Plus, Radeon, FirePro etc will remain. And, the quality of the products seems to be going up with every generation, all good.

Tonus said...

I'm not surprised that they're dropping the ATI brand, I figure that they're trying to consolidate things under a single brand. I don't know if it's a good or bad idea, the ATI brand has a good reputation these days. But I doubt that there will be much confusion. As far as I know, most of their sales are to OEMs, and a fair portion of their off-the-shelf sales are to geeks, who will know that AMD is ATI.

Anonymous said...

Just seeing if people are still reading:

AMD lowers Q3 outlook http://www.eetimes.com/electronics-news/4208754/AMD-lowers-outlook-semiconductor

"AMD announced that it expects revenue for the quarter ending Sept. 25, to be in the range of down 1-to-4 percent, as compared to revenue of $1.65 billion for the quarter ended June 26. Originally, AMD expected revenue to be up seasonally for the third quarter of 2010."

Intel also lowered its Q3 outlook 9about a month ago), however they lowered their outlook to flat/ slightly above Q2 (as opposed to AMD's down from Q2).

I wonder what impact this will have on GF as AMD is still loading most of those fabs.

Tonus said...

Any idea what GF's financials are like?

A Nonny Moose said...

http://articles.moneycentral.msn.com/news/article.aspx?Feed=FOOL&Date=20100926&ID=12085580&Symbol=AMD

Looks like the rumors about Oracle buying AMD may be true...

SPARKS said...

Who would have thought that AMD’s acquisition of ATI would ever bear fruit? I certainly didn’t, not then anyway.

AMD certainly couldn’t keep up with juggernaut INTC. It is abundantly clear they have become a niche player in processors, and are rapidly losing market share is the high end desktop and server arena.

However, the other player Nvidia (to which I personally have no great love for) has made a brilliant target for AMD. AMD has tiger by the tail starting with the 5xxx series and it looks like that trend will continue with the new 6xxx series. In fact, I cannot recall ATI/AMD having to successive generations of graphics cards trounce NVDA consecutively.

Perhaps Wreaktors dream of fusion will be realized, perhaps not the way he preferred it, but make no mistake powerful graphics with a mediocre processor may well indeed spell success for AMD in the long term.

This is where I see AMD headed, in graphics. In the mean time, who knows they may get lucky and come up with a real competitive CPU. But they would be well advised to continue full bore with their graphics processors, which they have done superbly.

I know, I have two of their graphics cards at 400 bucks apiece in my machine as I type.

After reading Charlie D.’s report, NVDA’s Some Dumb F..K is definitely having some sleepless nights. In fact it’s been a long time since AMD looked this good. If they are very lucky and very smart Wreaktor may have some vindication after all.

SPARKS

http://www.semiaccurate.com/2010/09/27/radeon-hd-67x0-series-specifications-leak/

SPARKS said...

Oh, as a side note regarding graphics, you were all correct in educating yours truly that Larrabee was not dead. (Yay)

FUDie has an interesting article regarding 22nM, no less. (I love the way they throw those numbers out there like, “The reason is simple, Intel plans to migrate to 22nm already in second half of 2011, while TSMC and GloFo won’t be ready by late 2012.”) 22nM, a piece of cake, right?

Anyway, if my beloved INTC can throw 5B at MacAfee, they can throw a billion or two at “Knight’s Corner”. This is a assuming the Mayan calendar does get the better of us all.

Hoo Ya.

SPARKS

http://www.fudzilla.com/graphics/item/20329-larrabee-comes-back-in-2012

Anonymous said...

There are another round rumors floating that IBM fab club is having issues with highK/MG on 32nm - thershold voltage shifts, oxide regrowth, thermal instablity(none of this would surprise me for a gate first process)

GF has dismissed the rumor. Samsung also denies this pointing to a low power 32nm process that was qualified.... of course the issues above would mainly be for HIGH performance processes, and the low power process would be less susceptible to these types of issues.

I'm interested to see (and would not be shocked) if AMD/GF's 32nm release is another "we're shipping lower speed parts because that's what customers want" type of launch (remember that ridiculous crap when K10 was having issues early on?)

What will also be interesting is not the launch of high K but the volume... a lot of the issues can have more of an impact on binsplits/yield as opposed to the process working/not working.

Tonus said...

Larry Ellison does sort of come off as the spiritual successor to Jerry Sanders, in some ways. Heh.

Anonymous said...

Things are so boring.

AMD and GF have folded a bit earlier then I thought.

The dream team in Pussyville NY has had NO record of doing anything right and gate first don't work. HighK and metal simply can't survivie the anneal temperatures required for the soure drains. Its fucking thermodynamics didn't IBM hire any people that got anything but an F?

Duh... I predict they will be very late.

Intel's biggest problem seems they can't design a low power CPU. Their designers have become so fat, lazy and can't design. They need some ARM like designers.

Both companies are f**cking up big time

SPARKS said...

"Their designers have become so fat, lazy and can't design."

Sure, i3, i5, i7, 32nM in full ramp, 22nM next year, fat dividends throughout the 'repression', bunch of fat lazy bastards.

I wonder what kinda juice ATOM will use at 22? Hmmph, who needs ARM?

SPARKS

InTheKnow said...

I wonder what kinda juice ATOM will use at 22?

We haven't even seen Atom at 32nm yet. Sure there have been a few presentations put out there, but no one can buy one yet. Until Intel rolls these out for sale, we won't really know what Atom looks like. In the mean time the smartphone/tablet market moves more firmly into the ARM camp.

I also get a kick out of ARM's A15 announcement.

This is the chip that is "going to allow ARM to move up the stack." And I agree it looks pretty good.

But what no one accounts for is how long it will take to get these to market. ARM's business model doesn't lead to rapid adoption of their newest designs. A15 will be out sometime in 2012 on 32/28nm.

By then Atom should be out on 22nm. So we don't even have a point of comparison between the two.

Axel said...

Last few days saw some rumors about Oracle possibly looking to buy AMD. CEO Dirk's words today hint that AMD would certainly consider an offer. If true, this means BIG implications for future AMD products. Unless Oracle are looking to get into the desktop & notebook business (unlikely), this would mean a drastic reduction in focus by Oracle-owned AMD on consumer products and more on the server & workstation space.

Axel said...

Oops:

Correct link

SPARKS said...

I knew ITK would beat me up on that comment, deservedly so.

It was once explained to me, in terms I could understand, that memory and logic is as different as a rocket and a jet engine. They produce thrust but in an entirely different way.

Companies show off cutting edge RAM, subsequently fall flat on their collective asses with logic. I called the whole shebang ‘The Memory Thing.’

The boys at INTC can build both rocket AND jet engines. Their new SSD lineup features Intel Micron Flash Technologies and the drives capacities are larger to boot, pun intended (bravo).

Imagine a 600 gig SSD! Hoo Ya.

Finally the irritating bottle neck between storage and processing may be coming to close. So goes the propeller by way of the jet engine and INTC’s new drives could be flying in the stratosphere………let’s hope the price isn’t up there too.

http://www.anandtech.com/show/3965/intels-3rd-generation-x25m-ssd-specs-revealed

SPARKS

Anonymous said...

Intel posted another solid quarter
11.1B vs 11B consensus
0,52/share vs 0.50/share earnings

Revenue up 18% year on year, up 3-4% from last quarter (AMD has guided flat to down and reports next week I think)

Gross margin projected at 67% for the 4th quarter (that is quite startling)

InTheKnow said...

Intel posted another solid quarter

And the analysts punish them for daring to do well during tough economic times. :(

A Nonny Moose said...

AMD revenue $1.62 billion, 2 percent sequential decrease and 16 percent increase year-over-year
Net loss $118 million, net loss per share $0.17, operating income $128 million
Non-GAAP(1,2) net income $108 million, EPS $0.15, operating income $144 million
Gross margin 46 percent

SPARKS said...

Moose, they know how to keep going, despite the losses.

“Cash, cash equivalents and marketable securities balance at the end of the quarter was $1.73 billion. The decrease from the second quarter was primarily due to the repurchase of $800 million aggregate principal amount of 6.00%

Convertible Senior Notes due 2015 offset by proceeds from our issuance of $500 million of 7.75% Senior Notes due 2020.

Approximately $780 million of the 6.00% Convertible Senior Notes remained outstanding as of September 25, 2010.”



However, there is money to be made. If you would have bought AMD on Aug 31 at 5.61 you could have done quite well tomorrow when they open at nearly 7 and a half!!!

You gotta have huge gilhonies though. I sure as hell don’t.

SPARKS

Anonymous said...

SO if I read the AMD press release right they bought back 800mil of convertible notes due in 2015 by selling 500mil in convertible notes in 2020 at a higher interest rate (7.75% VS 6.0%).... so just delaying the debt at a higher interest rate? (even with the difference in principle I'm not sure if that is a net win or loss)

Revenue was flat (very slightly down)... and people like the INQ were talking about Intel's report (they were actually up 3 or 4% sequentially). They also guided flat for Q4 (I think Intel's guidance was slightly up).

I think we are seeing a bit of the consumer/retail slowdown in the two sets of numbers... AMD tends to get hit harder in this area as they have better share (relative to the commercial space)

Quote of the day: "AMD's third quarter performance was highlighted by solid gross margin and a continued focus on profitability, despite weaker than expected consumer demand," (Meyer)

Good to see AMD still 'focused' on profitability. At what point does an analyst call them on what this means? Also solid gross margin of 46% (Q2 was 45%).... don't confuse 'solid' gross margin with improving gross margin.

A Nonny Moose said...

Read somewhere yesterday that AMD may have lost a bit more marketshare in server, due to Magny Cours being a socket change (from socket F). Or IT buyers might just be waiting for Bulldozer and bypassing MC.

The situation might be reversed when Bulldozer (drop-in upgrade) is competing with Sandy Bridge server (requiring a socket change).

SPARKS said...

Moose, how many times did I rant about an AMD socket change in the past? I knew they’d take a hit if they did.

Surely, every AMD fanboy upgraded on the cheap with AMD’s simple 250 buck drop in replacement since 2007----- if he could. Black Edition, oooh!

However, if someone is going to spring for a new motherboard, CPU and memory, well that’s a different story. This is where the real competition begins, not with a cheap drop in, but on a platform to platform basis.

With INTC’s i3, i5 and i7’s there’s simply no contest. No one builds an entirely new machine without seriously considering the best bang for the buck.

AMD held out as long as they could. Think of all those 50 and 100 dollar dogs that will be sitting on the shelves rendered obsolete overnight. (Remember 939 t0 940?)


http://www.newegg.com/Store/SubCategory.aspx?SubCategory=22&name=AMD-Motherboards


Watch the AMD fanboys bail en mass.

I made the prediction that AMD will be relegated to the Whitebox/Budget gulag and I stand by it.

SPARKS

SPARKS said...

Well G, Andrew Como is our new Gov.

I think I'll cross a phase in a 6000A 277/480 service tomorrow morning.

We're doomed. Well at least he's off the 'get' INTC bandwagon.

SPARKS

Anonymous said...

Don't worry, he's probably mapping out how he can run for prez in 2016.

That means he can only do damage over a ~4 year period as the exploratory committees and campaigning will likely start in 2014, maybe even in 2013.

Depending on the political climate he'll probably take a couple of more runs at banks and large companies... oil doesn't overlap much with NY so sadly he can't take on that 'villain' to score points.

The beauty of being a Democrat in NY is that once you win the primary, the general election is an afterthought and you can start mapping out the run for the next higher office... I'd expect either Schumer or Cuomo to start positioning themselves post Obama in 2016 (whether or not Obama wins in 2012)

SPARKS said...

“The beauty of being a Democrat in NY is that once you win the primary, the general election is an afterthought.”

No doubt, so much so it makes one wonder what the hell the state Republicans were thinking when they put up that maniac Paladino to run against him! It’s as if they handed him the Governors seat on a platter.

Como for President in 2012? Nah.

Strike One- New York Liberal.

Strike Two- NO ONE will swing the minority vote from Obama in 2012, he’s a shoe in.

Strike Three-Americans of Italian descent don’t fair to well nationwide. Think the Corleone family, and the Soprano’s. Even Rudy didn’t have a prayer. (In the interest of discloser I am of Italian descent).

Taking a run at the banks, you say!?!? The banks were underneath that 600B piñata the Fed held up for grabs yesterday swinging away like a bunch of drunken teenagers who sneaked a couple of beers to the party.

No one can touch the damned banks, not now anyway. They get free B-plus loans while most get $500 secured credit card loans to the tune of 20%. Those of us with some money, however, are yielding 1 to 1.5% on treasury notes. Why should the banks go to us, when they’ve got the Fed feeding them carte blanche right from the nipple?

These stimulus packages remind me of a doctor who’s standing over a day old corpse, paddles in hand, screaming, “See, it moved!!!”

Obama in 2012, the DARLING of the banking industry.

SPARKS

Hチェッカー said...

自分のエロさってどのぐらいなんだろう?と思ったらHチェッカー!年始の期間限定で特別に診断結果に合わせて、異性を紹介するおまけつき!自分と同じぐらいHな異性と知り合えるのですぐに仲良くなれる!

Anonymous said...

Bake at 350 for 25 minuteѕ or untіl toothpicκ inserteԁ in сenter comes out сlean.
Thiѕ аnnual рlаnt cаn be gгown in аn contаiner, anԁ wіll уield betwеen 1-2 сups of freѕh
bаѕil. I trіed ԁozеns οf diffeгеnt гeсipeѕ for pizza cruѕt, and none of them were
satiѕfactory.

Here іѕ my ωеb blog :: pizza stone
my web page - pizza pan alexandria

Anonymous said...

I'd like to thank you for the efforts you have put in writing this site. I am hoping to view the same high-grade blog posts by you later on as well. In fact, your creative writing abilities has motivated me to get my own, personal blog now ;)

Feel free to visit my web-site: adventistmembers.com
Also see my webpage :: painting techniques in bathroom

Anonymous said...

She got the ideа frοm cooking рioneer Barbага Κerr.
Sіtting down there on уοur counter,
it rarely sееms to bе capable of this soгt of miraсlеs, but thе νеry firѕt timers
and the ѕеаsoned cooks aliκe will saѵoг incoгpοrating that standard flavoг to thеіr favoritе dishes.
Thе Emancipation of Mimi is the tеnth studiο album by Amerіcan singer Mariah
Сarey.

my blog :: pizza stones for sale sydney

Anonymous said...

Wow thаt ωas oԁd. I just wrοte аn very long comment but after
I clіcκеd submit my comment didn't show up. Grrrr... well I'm
not wгiting аll thаt οver again. Anуhoω, just wantеd to say
fantastіc blog!

My homepаge :: zombie-Killers.Org

Anonymous said...

It is ԁеsigned up from different graԁes of minοr combinatіon that have been bеfoгehand coаted in ѵeгу hot bitumеn that functiοns as a binder when mixeԁ tоtаlly with warm asphalt.
- Centгe tunnеl chimney duct tо maximise effective vеry hot aіrflow.
As a matter of ρoіnt it is this sort of а hugely-concentrated fоoԁ
that, except taken in comρletely ѵery ѕmаll
quantities, it is liable tο upset weаk digestiοns.


My ωebsite Old Stone Oven

Anonymous said...

After I initіаlly сommented I
appear to haνe clickeԁ on the -Notify me when new cοmmentѕ are
adԁeԁ- cheсκbox and now whenever a comment iѕ addeԁ
I receive four еmailѕ with the same сοmment.
Perhаps thеre is a way you can remove
mе from that ѕеrviсe?
Kudoѕ!

Feel free to viѕit my weblog :: Augen lasern

Anonymous said...

It's actually very difficult in this active life to listen news on TV, therefore I just use web for that reason, and obtain the hottest news.

my web site: augenoperation

Anonymous said...

I don't know if it's just me οr if everyone еlse eхperiencing issueѕ
with your ωebsіte. It aрρeaгs as if ѕomе of the written teхt in yοur content are running off
the sсreen. Cаn somebody еlsе
pleaѕe provide feedbacκ and let me know if this is happening to
them too? Thіs may be a ρrοblem with
my web broωseг because I've had this happen before. Thank you

My blog post Chemietoilette

Anonymous said...

ӏ would not ρut it in the freеzer for a long timе, you can
put it in the freezer for a couρle of houгs but I would nоt put it in theгe for a couple of weeks.
All yοu have to ԁο in order
to obtain thesе cоuponѕ іs to gо to
the internet and search for the one that suits you needs thеn just
ρгint them. I'm sure you'll find our reviews to be honest and аccurаte.


Mу web pаge pizza pan ashtabula ohio

Anonymous said...

Caгеfully рeгmіt go and then set thе
spider upsіde down (with its legs in the aіr) to dry.
At mіnіmum 51 % have to be com, although mostly up to seventy five%
іs made use of. Stir in onionѕ, сelery, chopped
yellоw bell pepper аnd cooκ
dinner bеfoгe tenԁer.

Ѕtop by my web-site old stone oven

Anonymous said...

It iѕ actually a chef's expression for 'a put for just about
everything and anуthing аnd everything in its plaсe'. Get ready the spot just where you are heading to be doing the colorng. Fabric markers have even been chosen and can be advantageous to contact-up the areas on the sneakers whereby the colour did not choose (the seams primarily).

Stop by my webpage how to make pizza using a pizza stone

Anonymous said...

I lіke it when іnԁiνiduals comе
together аnԁ share оpinions. Great website, сontinue thе
goοd wοrk!

Feel free to vіsit my blog :: Augen Lasern