Saturday, October 28, 2006

Intel's Bluff (Q3 06 Results)

It's taken me a little time to go over Intel's finances in detail. However, far from being the glowing outlook that nearly everyone is saying, Intel's financial position is much more precarious. The good third quarter results were actually only a temporary fix created by slashing stock buyback. Unfortunately, even following the buyback patterns from two years ago, this will still leave Intel about $2.5 Billion short at the end of the year. The best that Intel can hope for is that it has a much better fourth quarter than even their best projections.

The overly positive outlook for Intel seems to be based on three assumptions:

  1. Intel is regaining share.
  2. Intel is making more money.
  3. Intel is back to playing its own game.

I've seen some very poor attempts to give Intel credit for a supposed increase in volume share by using a bit of statistical sleight of hand. The statistics for the last three quarters for microprocessor volume are:

Intel - 74.3%, 72.9%, 76%
AMD - 21.1%, 21.6%, 23%
VIA - 4.6%, 5.5%, 1%

Based on these percentages, the naiive claim is that Intel gained 3.1% in the third quarter while AMD only gained 1.4%. This would seem to suggest that Intel gained twice as fast as AMD in the third quarter and, presumably, this would be an indication that Intel is taking back share. Unfortuately, this distorted view is caused by the interference from VIA. VIA only sells at the very low end of the X86 market. Adding their share in with AMD's and Intel's is a bit like adding bicycles to motorcycle sales. We can get a clearer picture by removing VIA from the volume statistics and just showing the volume share between AMD and Intel:

Intel - 77.9%, 77.1%, 76.8%
AMD - 22.1%, 22.9%, 23.2%

The undistorted statistics tell a very different story. We can see that Intel has been slowly losing share to AMD for the last three quarters; quarter three was no exception. Microprocessor revenue share is similar:

Intel - 82.8%, 81.9%, 81.7%
AMD - 17.2%, 18.1%, 18.3%

Clearly, Intel is not regaining share in either microprocessor revenue or volume. Similar distorted claims are based on operating income and cash:

Intel
Q2 - Revenue - 8009, Op. Inc. - 1072, 13.4%
Q3 - Revenue - 8739, Op. Inc. - 1374, 15.7%

The operating income as a percentage of revenue appears to be up and Intel boosted its cash by nearly $700 Million in the third quarter. This looks very positive for Intel. However, there is a problem. People who buy Intel stock rely on Intel's stock buybacks to increase demand. In 2004 Intel spent $7.5 Billion on stock buyback and spent $10.6 Billion in 2005. Based on the lower 2004 pattern Intel should spend $7.7 Billion this year. However, Intel has only spent $4.4 Billion so far and should have spent about $1.9 Billion in the third quarter. Instead, Intel only spent $500 Million.

It is clear then that Intel boosted its third quarter numbers by slashing its stock buyback. This makes the third quarter numbers look good but puts Intel $1.4 Billion behind in addition to the $1.9 Billion it would need to spend in the fourth quarter. However, if Intel spends the same as it did this quarter it will end up $2.5 Billion short which will cause its stock to plunge. If it pulls the cash back out that it just put in this would end up only $1.2 Billion short. This is a tough position for Intel. If it doesn't spend any cash then its stock buyback will drop below the 2004 levels and will only be a little ahead of the 2003 levels. This would be a 50% reduction from the 2005 levels and something the stock market surely would not overlook. This could easily cause Intel's stock to slump when the end of year results are released(which will be sharply down from last year). Unfortunately, Intel will not have the money to spend so it will have to choose between spending cash and risking a drop in stock price. From this we can see that Intel's money position is currently worse than it was even in 2004. Intel has not yet recovered in terms of money.

The final assumption is that Intel had to play AMD's game while AMD was ahead in processor performance and now that Core 2 Duo is ahead Intel can play its own game. Well, let's assume for a moment that this is true, that C2D's superior performance is creating lots of demand for Intel and eroding AMD's demand. This could cause AMD to drop its prices even lower to avoid losing too much volume share and this would hurt AMD's margins even more.

The problem is that AMD hasn't lost in volume share; Intel actually lost a little. If the assumption about demand were true then this wouldn't fit with the facts at all.We know that AMD was capacity limited in the third quarter. If demand were strong then Intel should have had no trouble picking up customers beyond what AMD was able to supply. But, this didn't happen. Secondly, the fact that Intel's margins were lower than AMD's and will stay lower in the fourth quarter disproves the notion that AMD has to compete by dropping prices. A similar idea was that Intel was going to get rid of its overstock by selling it at reduced prices and this was supposed to undercut AMD and cause further loss of margin. However, AMD's margins didn't drop all that much and Intel ended up having to take a loss on $100 Million worth of overstock. Apparently, this was only part of the overstock and Intel may have additional losses in the fourth quarter.

The fact that Intel was unable to gain volume share while AMD was capacity limited disproves the idea of extra demand for Intel chips. The fact that Intel didn't gain at all in terms of revenue share also disproves the idea that Intel's chips are more valued. Intel finds itself faced with a competitor who is serious about growing capacity. The only way of stopping this would be to somehow reduce demand for AMD chips and this hasn't happened. Intel seems to be overlooking the fact that having two FABs plus second sourcing from Chartered makes AMD a very reliable supplier; Intel has lost this advantage. Intel has been the strongest player by far in terms of integrated graphics for both desktops and mobile. However, now that AMD has ATI Intel faces a whole new level of competition in these markets.

AMD is currently taking mobile share from Intel. Many have assumed that Merom, the mobile version of Conroe would easily hold this market for Intel. However the fact that Intel will keep making the lower performing Yonah until the end of 2007 suggests otherwise. It appears that Merom is having more difficulty with power consumption than the older Core Duo. AMD's Turion should compete with this very well as it moves to 65nm. This seems to have been a strategic mistake for Intel and not one that it can fix soon. Faced with an inability to manage multiple processor projects, Intel lumped everything together with Core 2 Duo. This chip tries to do the job of server, desktop, and mobile. Apparently, it isn't doing mobile as well as it was expected. This unfortunately occurred at the same time that AMD decided to split mobile off into a separate family. This seems to be what will give AMD the ability to take share in mobile away from Intel. Intel surely realizes its mistake and is probably frantically trying to reform another all mobile team as it had with Pentium M. However, this is likely to set Intel back into 2008 and it can lose a lot of share during that time.

The one size fits all strategy with C2D is also not working as well as it could with servers. Woodcrest doesn't do 4-way and Intel won't have a 4-way until Q3 07. This means that during this time Intel has to compete with the outdated Cedar Mill based Xeons. The result has been that Intel has not been able to take back any of the 4-way and higher share that it lost to AMD both from Xeon and Itanium. Given AMD's recent wins in supercomputers with Opteron it seems that AMD will remain strong in the top end of servers well into 2009. Although Intel has made some gains in 2-way servers it will soon be faced with both lower power 65nm based Opterons and a native quad core design with K8L. This is a fairly grim prospect for Intel as it won't give them much time to gain any share and even then only at the bottom. Further, Intel appears to be behind in base architecture and it doesn't appear that they will catch AMD until 2009. AMD has a good chance of reversing any server losses during that time and taking additional server share.

The sole bright spot for Intel is the desktop. Or maybe it isn't. This depends on how much of Intel's third quarter volume came from moving lower priced P4 stock. It could be the case that Intel's rise in desktop volume share was only a one quarter bump and will drop again in the fourth quarter or even the first quarter of next year after it moves more overstock in the fourth quarter. At this point though there isn't any way to tell. It seems reasonable though that, with increasing pressure from AMD within the corporate market with its Stable Image Platform and integrated graphics from ATI, AMD is not likely to lose very much share. Its deals with the two highest retailers in China and its deal with Dell also suggest that any volume loss on the desktop will be limited. Some have tried to argue that Dell is rapidly losing ground and therefore won't help AMD. However, this seems to be more of a sour grapes argument than anything else. Looking over Dell's financials I was unable to see more than a drop of about 4% and this is hardly proof of any downhill slide for Dell. Intel dropped much more than this during 2006.

Although it has been suggested that AMD's current capacity constraints will cause it to lose customers this logic too does not appear very sound. AMD's capacity will increase rapidly next year so Intel can't count on capacity constraints for AMD unless the market grows much more rapidly than it has. Then too, Intel still would have to turn these limitations into demand for its own products as it was unable to do in the third quarter.

Intel is also unlikely to improve its popularity with vendors. Even though AMD now has ATI it is quite willing to let its customers use other brands of chipsets if they prefer. Intel in contrast, is much more insistent about pushing its own products as it has with its exclusive chipset policy for Centrino.

I have been unable to find any evidence in the third quarter reports that Intel's prospects are better than AMD's. Intel has been unable to create demand or decrease demand for AMD. Intel will find itself in a defensive position trying not to lose share in mobile and corporate accounts while trying to push a single chip design that is not fully adequate for either server or mobile use. It has to do this against a tough competitor who is putting huge amounts of its earnings into capital expenditures to grow its capacity. On top of this, Intel will find itself with a very disappointing 2006 earnings report and enormous pressure on stock buyback with money that it won't have. It is likely to be faced with the prospect of either spending down its cash even more or watching its stock price plunge. From all of these disadvantages it is clear that Intel is not playing its own game.

94 comments:

Erlindo said...

Thanks Scientia for your valuable forecast.

AMD and Intel are punching each other with whatever they have, but I do hope AMD to do much better next year with their upcoming Barcelona core.

What has me worried is the initial speed of the K8L cores: 2.5/2.9GHz vs intel's upcoming 3.5GHz quad cores. I do hope AMD's numbers to be conservative and intel's numbers to be part of their marketing BS. :)

Another question: would it be enough for AMD to acquire 35% market share to have an even fight with intel or do you think that AMD would need more than that to have their market share assured in the long term?

Thansk in advance and keep on with the "good blogging" ;)

ashenman said...

I think both companies will be performing some adjusting on those next gen quad core numbers. I was hoping someone would finally show volume percentages that didn't involve Via. This puts AMD in a much better light than I expected. In reality, I thought the two companies would be flat, but apparently AMD is still gaining.

Although this does look great, I still stand by the prediction I just made on 180's blog that q4 will be flat for both AMD and Intel. I think demand spikes for the season will be met proportionally by increased output from their manufacturing. Intel will also be meeting the seasonal increase with an increased ratio of core processors to netburst, which seems to be dying a bit faster than most people think.

I think the mixes of mobile to desktop will change alot though, but I wont bother to elaborate there, because there are too many variables to even guess how it will end up.

If anyone pays attention to "Chips and Salsa" you'll notice he apparently thinks that Intel releasing quad core server chips 1 qtr earlier will destroy AMD's 4p segment. While 180 and core2dude also site that woodcrest moved to 40% of the market already, they do not mention that opteron 2p basically stayed flat, meaning that Intel was unable to capitalize on woodcrest because it has too much production based on netburst. This wont change much by q1 07. Intel's woodcrest production looks like it might reach high enough numbers to replace its netburst production, but opteron will still have pretty much everything else as prices drop and offers from dell become available.

I also think it's worthy to note that part of Dell's recent lambasting by analysts has been because its throne as largest volume distributor was taken by HP. However, that's not because Dell was shrinking, but because HP was growing, and it was using AMD desktops and servers. Dell is being extra careful right now, and it's not about to go making mistakes that could potentially anger its newest and potentially most important cpu supplier (if anything, at least in dell's eyes).

Scientia from AMDZone said...

I just get tired of all the biased talk. Sharikou keeps talking about Intel's going bankrupt while Sharikou180 talks about Dell's going bankrupt. There is no indication that Intel is going bankrupt but I also can't find any indication that Dell is hurting that bad either. I get tired of seeing people trying to spin the numbers to magically make Intel have a better quarter. And, it irritates me more because this Pollyanna view of Intel is also common among supposedly professional analysts.

It could be a problem if Intel really can get its clocks up that high and AMD stays low. I'm just not so confident in the supposed big jump that Intel will get from 45nm. I'm also not quite so pessimistic about AMD. We'll see.

The simple fact is that Intel is moving back up but it is still barely above 2004 revenues. However, Intel was able to spend $7.5 Billion in 2004 on stock buyback but it doesn't appear that Intel can afford that this year unless they spend down their cash.

AMD is well ahead of 2005 revenues or about double what it made in 2004.

I think Intel will do better and will keep growing. I just don't have the confidence that Intel will take back share from AMD as everyone has been suggesting. I honestly think it is more likely that AMD will grow a bit faster.

I'm thinking that Intel's monopoly will start breaking in 2008. Basically, I think AMD will be doing ok with 25% revenue share but good with 30%. This means that AMD needs to grow about 50%. If AMD grows as fast as it did then this would be about 14 months. However, including the normal expansion of the market this would be closer to 18 months.

Scientia from AMDZone said...

If anyone pays attention to "Chips and Salsa" you'll notice he apparently thinks that Intel releasing quad core server chips 1 qtr earlier will destroy AMD's 4p segment.

A couple months ago I looked over chips and salsa and was disappointed. I just went over there and read the last six posts. Core2Dude doesn't seem to be very familiar with the markets or processor technology in general. He just seems to be be cherry picking any factoid that seems to support his biases. But, he doesn't seem to be able to fit any of this information into a bigger picture.

ashenman said...

"This means that AMD needs to grow about 50%. If AMD grows as fast as it did then this would be about 14 months. However, including the normal expansion of the market this would be closer to 18 months."

When is "then"? I'm assuming that AMD will grow a bit faster than this, but only by a couple months, just because I don't think we'll see normal growing trends for AMD or Intel, and that they will start to fluctuate in marketshare by substantial amounts past q3. The reason being is both will have highly competitive products and both will have healthy amounts of the market, and both will hopefully (both in Intel and AMD's case) have stable and effective business images by then.

Anonymous said...

The sole bright spot for Intel is the desktop. Or maybe it isn't.

You're right, there is absolutely nothing going on for Intel;)

I didn't know that stock buyback was factored into earnings..Even so, if Intel 'only' spent $500M this quarter, this gives them good room to not miss future earnings.

I don't think Turion X2's tech is what is selling it;)

Good luck to AMD, with a new processs, new core, and that huge die come Q207;).

Lose customers? In what sense? During Intel's shortage, some looked elsewhere, but that doesn't necessarily mean that they've gone AMD only.

What kind of popularity do they need with vendors? Things like Alienware being AMD exclusive at Digital Life?

http://www.gamepc.com/labs/view_content.asp?id=c2dm&page=14
Core 2 is better than Core 1. What exactly about it is so lacking that would require a new design?

Intel's dissapointing 2006 report will not be a surprise to anyone:)

PS, please explain this game that AMD played when they were ahead:D

Darth Solarion said...

Stock buyback is one of the many accounting tricks one uses to inflate numbers.. quite a few companies do that...

ashenman said...

While the 230mm^2 die for barcelona is pretty big, it does not mean that this processor will inevitably be hot, or that it will be slow or expensive. Part of the reason Via is able to make their systems so cheap is because they produce on larger dies, and simply refine their production so much that they can mitigate the likelihood of defects that increases in larger wafers with better and safer processes. I think that's part of why AMD is taking its time with 65 nm and working so closely with IBM. I also think it's why AMD is going to ramp 45 nm faster, because their toolsets are designed for it, and thus their enhancements will be directly applicable to that process.

Scientia from AMDZone said...

Red, are you a native English speaker?

I'm asking this because you seem to essentially understand English but you seem to miss the subtleties in most of my articles.

Scientia from AMDZone said...

The whole point about stock buyback is that in 2004 Intel could afford to spend $7.5 Billion on it while adding $900 Million to cash.

Yet, in 2006 with bigger revenues Intel is behind $1.4 Billion in stock buyback. To match the 2004 levels Intel would need:
$200 Million - cash
$3.3 Billion - stock buyback
Total - $3.5 Billion.

Intel will not have this much in the fourth quarter. This means that Intel will be doing worse in 2006 than it did in 2004 even though it will have more revenue.

Anonymous said...

http://techreport.com/onearticle.x/10898
300mm for 1 die of Barcelona, 286mm for Kentsfield(2 Conroes). They won't be hot like nothing we've ever seen, but still, at the max TDP, and not able to nudge past 3.0. PS, please tell me more of Via? and its huge dies:) According to Wikipedia, they seem extremely small.

AMD and 45 nm should be interesting, considering that they're still trying to get 65 nm as fast as they can. Any dates on 45 nm? I see TSMC going for test production of low power 45 nm in Q307..
---
Do you lack courtesy? Questioning my comprehension of English? God forbid I'm Hispanic or Asian. Which I am, thanks for the offensiveness! It is a bit hard to understand your articles, with this one at 2000 words!

Whoops:p I don't think I suggested that Intel is going to do hot this quarter, but 'only' spending $500M shows how much leeway they have.
"Intel's dividend and stock buyback programs could be affected by changes in its capital spending programs, changes in its cash flows and changes in the tax laws, as well as by the level and timing of acquisition and investment activity."
There, they have no desire to continue that goal.

ashenman said...

Sorry, 291, or about 300mm^2. Via's processors use older mfg processes at 130 nm. This makes them relatively large compared to what they could be. I'm mainly operating off of hearsay though, so I guess it's not a valid point.

45 nm will take 18 months or less to ramp. I'm assuming they'll start ramping as soon as 65 nm is done.

Core 2's problem is it will be sharing 4 cores across one cache in their unified core design will be highly problematic, as it will lead to much cache thrashing. Kentsfield's problem will exist when one core wants to talk to the other on the other die, or when both cores on one die want to talk to the other. While that doesn't happen much now, it will happen as programs begin to take advantage of multiple cores, as they basically don't right now, even if they say they're optimized.

Scientia from AMDZone said...

Red, I'm sure you can understand basic English but I tend to write at a high level and you do not seem to comprehend all of it.

Also, when you make comments here, your writing tends to be a bit jumbled. I would suggest that you avoid both sarcasm and rhetoric. Sometimes it is difficult to tell when you are being sarcastic and when you are being serious. It can be difficult to tell when you are being rhetorical and when you are actually asking a question. I would also suggest that you avoid slang terms like "hot".

You're right, there is absolutely nothing going on for Intel;)

Presumably, this is sarcasm. Unfortunately, it also shows that you didn't really understand my article. What I actually said was that Intel might be doing well on the desktop if the volume increases this quarter were not due to discount P4 sales. However, there is no way to tell since desktop sales are not divided into individual types.

I didn't know that stock buyback was factored into earnings.

Stock buyback is not counted as income and is deducted before taxes. However, it still has to be paid by Intel's revenues.

Even so, if Intel 'only' spent $500M this quarter, this gives them good room to not miss future earnings.

No. This makes it tougher for them to avoid a stock plunge.

I don't think Turion X2's tech is what is selling it;)

Presumably here you are trying to be subtle. The problem is that you end up saying nothing and your remarks are just ignored.

Good luck to AMD, with a new processs, new core, and that huge die come Q207;).

I suppose you are suggesting that AMD will have trouble in 2007 with production. However, this is incorrect. The 65nm process has been tested since November 2005. The new cores will have been tested for about a year before production. Only the quad core dies are large and these are the smallest production volume. I think most people are more interested in the dual core version of K8L which will be a fairly large upgrade to the current K8's.

Scientia from AMDZone said...

45 nm will take 18 months or less to ramp. I'm assuming they'll start ramping as soon as 65 nm is done.

I believe 45nm demonstration has already been done. 45nm testing would be about mid 2007. 45nm test production begins at the start of 2008. This means that production release will be about mid 2008. FAB 38 and FAB 36 will both still be ramping when 45nm testing begins. 90nm production will end perhaps one or two months before 45nm release.

Anonymous said...

What about my statements would indicate that I 'don't seem to comprehend it all'?

Scientia: You have such a good grasp of the English language. I'm amazed how you cannot see the significance of Intel cutting back on stock buyback while they are still in the midst of restructuring. This would obviously indicate that you are not a native English speaker, as only foreigners, with even the greatest grasp of English, would never be able to understand this article.

"The sole bright spot for Intel is the desktop. Or maybe it isn't."
Well if they're only shining star is not shining, one would be able to conclude from your article that you see nothing shining from Intel?

..$500M leeway from last quarter. Their stock buyback alone is greater than AMD's earnings. I don't see how they're in danger of a collapsed stock, not like there is much to collapse to.

We don't see them exactly ramping up the clocks on 65 nm. They will have less than a year from now to get Barcelona ready. A0 doesn't look promising and they could still be working on kinks til then for all we know.

ashenman said...

Red, it doesn't matter if their buyback is greater than AMD's earnings. If anything, this argument works against you, as it shows just how inflated buyback makes things. Also, it shows how much this taxes Intel's financials. Not only that, but the tech industry normally swings pretty fast on the side of manufacturers like Intel, AMD, and most notably, sony. This 500 million dollars can mean alot for Intel, and that's the problem.

AMD has historically been very relatively modest about what it plans for its parts this far out. They also normally start slower than their last gen chips and quickly ramp up to higher speeds very quickly when distribution becomes even.

Scientia, while Red does seem to lack distinct comprehension of certain specifics of your arguments, that doesn't mean he speaks another language. Even then, it's rude to ask him that. I think you're being a bit presumptuous in both the complexity of your statements and the origin of Red's inadequacies. Your posts are really long, and sometimes if you want to read it in a reasonable time frame you can have a hard time catching everything. I think you need to take a step back and recognize what differentiates you from MMM, Core2, and Sharikou. You're still doing a good job, but you're slumping a bit, and need to either post less or post more effectively. I also think you need to find a different way to set up posting, as this blog seems to be suffering in terms of readership and discussion.

Red, you do miss out on some of Scientia's points, to the degree that I sometimes wonder if you read them at all. I think you need to make sure you read his posts at least twice to catch everything. Also, if you write with lots of nuance or rhetoric, point it out, because if anyone skims the main post, they really skim the comments, and don't catch any of that that's not specifically written out.

ashenman said...

Sorry, by posting I mean commenting. However, I think you do need to refine your posts a bit to be more concise. Is there a way to have anonymous comments require acceptance on the bloggers end while letting non-anonymous ones through. As much as I enjoy talking with you and Red, I think others are missing out and have some good questions they could be asking.

Anonymous said...

Besides bandwidth, and I am just throwing this out there so don't slam me, but what if Intel was to use its manufacturing muscle and produce a lot of quad cores?

AMD's limited capacity would be hammered, especially if Intel is really building those 3 45nm FABS.

Those FABS plus, (isn't it 3?) 65nm FABS would be very powerful.

AMD does not have the capcaity to do this and the end result would be (just speculating here), Intel quad core $500, or AMD dual core $500 the death of highend desktop and 2P servers for AMD.

I have asked this before (not here), but isn't 2P about 80% of the server segment, and wouldn't this crush AMD?

Just another view,
Thanks.

Scientia from AMDZone said...

Scientia, while Red does seem to lack distinct comprehension of certain specifics of your arguments, that doesn't mean he speaks another language.

No, it doesn't. My question was based on observations of Red's posts for the past 3 months. Not being a native English speaker would explain a lot.

I think you need to take a step back and recognize what differentiates you from MMM, Core2, and Sharikou.

What?

You're still doing a good job, but you're slumping a bit, and need to either post less or post more effectively.

What do you mean?

I also think you need to find a different way to set up posting, as this blog seems to be suffering in terms of readership and discussion.

123 people viewed the blog today. I typically get around 50 a day. Apparently, not being able to comment anonymously discourages people from posting. Do you think I should allow anonymous posting again?

Anonymous said...

Scientia said...

"Do you think I should allow anonymous posting again?"

I know your not asking me, but...

No, that leads to another Sharikou site, no debate, just flaming.

Scientia from AMDZone said...

what if Intel was to use its manufacturing muscle and produce a lot of quad cores?

Apparently, Intel can't do this yet. The volume should be good by Q2 07 though. Intel couldn't make a lot of quads without cutting its margins. Or do you mean only server?

AMD's limited capacity would be hammered, especially if Intel is really building those 3 45nm FABS.

FAB 32, FAB 28, and D1D. However, D1D is not a production FAB.

Those FABS plus, (isn't it 3?) 65nm FABS would be very powerful.

D1C, FAB 24-2, FAB 12, and FAB 11X. I guess that would be 4.

AMD does not have the capcaity to do this and the end result would be (just speculating here), Intel quad core $500, or AMD dual core $500 the death of highend desktop and 2P servers for AMD.

It would cost Intel money but I suppose it would be possible to flood the server market with Clovertowns. I guess I would have to give this some thought to see if something would prevent it.

I have asked this before (not here), but isn't 2P about 80% of the server segment, and wouldn't this crush AMD?

Would this also undermine Intel's desktop market?

pointer said...

Well essay, but with wrong logic to start with. it is just like saying 1+1 should be equal to 3, that's why i told you 1+2!=3, 3+4!=7, ... so on and so forth.

some thing for you to think:
Adding their share in with AMD's and Intel's is a bit like adding bicycles to motorcycle sales.
you see only bicycle and motocycle? what about the car driver, sport car driver, airplain pilot out there? no matter how you spin, it is still x86 market. AMD (and Intel) would love to eat into whatever market it can go. If you seriously think those 'cheap' segment doesn't count. So you would not count in whatever Geode sold, would you? (it is going to be quite high volumn)

you also assume that Intel grap the most bicycle rider convert. it can be really amd grap Via's while intel grap AMD's. allow me to learn from you and sharikou, that would means that Intel has real 76%, and AMD has 19.4% after deducting the bicycle that they get.

And finally, during last quarter, of any blog comment that you posted ... you would(did) say intel share drop by 1.4% instead of the 0.8% of your 'new' calculation right?

Anonymous said...

Scientia from AMDZone said...

"Would this also undermine Intel's desktop market?"

It would probably have an effect, but if they lower their prices for dual cores, and make up for it with the quads... I am no economics major, so like I said, I am just throwing this out there, but would you buy a dual core for the same price you could buy a quad core?

The more I think about it, this would be more for 2P, and not desktop, but the same question 4 or 2?

Scientia from AMDZone said...

you also assume that Intel grap the most bicycle rider convert.

No. I'm saying that neither AMD nor Intel compete at the low prices where VIA sells its processors. However, AMD and Intel do compete very heavily with each other.

allow me to learn from you and sharikou

If you honestly believe that my articles are like Sharikou's then there is no reason for you to comment here.

that would means that Intel has real 76%, and AMD has 19.4% after deducting the bicycle that they get.

First of all they didn't get anything from VIA which is why I exclude it from the calculations. Secondly, your numbers are completely wrong.

with VIA
Intel - 76%
AMD - 23%

without VIA
Intel - 76.8%
AMD - 23.2%

And finally, during last quarter, of any blog comment that you posted ... you would(did) say intel share drop by 1.4% instead of the 0.8% of your 'new' calculation right?

Yes, I think it is more accurate to say that Intel only lost 0.8% of volume share to AMD in the 2nd quarter. This also means that AMD only gained 0.3% this quarter instead of 1.4%.

pointer said...

If you honestly believe that my articles are like Sharikou's then there is no reason for you to comment here.

sorry for the insult :) I sincerely apologize for that. :P

Scientia from AMDZone said...

No. That wasn't what you said before. The reason why Intel can't just start producing lots of quads is because it would drive down their ASP and therefore their margin would suffer.

However, if we restrict the argument to servers as you were saying before then it's a little different. You could argue that Intel could lower server ASP as long as it stayed within the boundaries of increased share. In other words, Intel wouldn't want to convert a high value server share to a slightly large share with much less value. This would prevent Intel's just dumping quad Xeons but not necessarily prevent more competitive pricing. I guess we'll have to see how the pricing goes in Q3 07.

pointer said...

First of all they didn't get anything from VIA which is why I exclude it from the calculations. Secondly, your numbers are completely wrong.

Q3 is normally a better quarter compare to Q2. If AMD and Intel really didn't grap anything from VIA, who did? Or you have any ifo that indicate spotanenous sales drop of VIA without VIA CPU replaced by either AMD/Intel? In another word means those VIA customer is not buying anything by about 6X.

ashenman said...

I guess with the comment on finding a new way to manage comments, I was just feeling a little constrained with having only me, you, erlindo, and red commenting.

With the taking more time between posts comment, I actually meant to write comments, because I thought the argument on semantics was getting a bit drawn out.

On the stepping back and taking a look at the differences between... comment, I was just saying that all those writers tend to judge their commentors beforehand, thus resulting in massive flaming and burning.

Basically, I was mainly referring to comments, but was really really tired because I'd just had a polisci test and and done a bunch of engineering homework, and had a hard time forming statements and sentences and whatnot, so I really apologize for the ambiguity and poorly made (I guess you'd call them) arguments.

A good way to test the Via argument, is to build a Via system, and then try to build something with an offering by Intel or AMD. And the result would be that you couldn't with Intel, and that you could if you had a contract with AMD. Via makes processors that are specialized in embedded kiosk like systems, in plcs (I think that's what they called, but basically specialized yet general usage manufacturing computers) and mini-itx systems. 180 pointed me to Intel's embedded site, which basically offers their general usage processors, which would in no way be ideal for embedded. Mini-itx is the same. While there are Mini-itx boards supporting amd and Intel processors, they're hard to find, very expensive, and don't have as many features, which is what matters most when building a mini-itx system. Via's bump was caused by a new deal to supply a laptop builder that needed super light and super long life batteries while keeping the laptops very cheap (under $500). I also think the deal was made because of Via's onboard cryptography (they still do that right?) for business systems. This was a prebuild scenario, much like AMD's gain in laptop share from Dell's prebuilding their sub $500 laptop. Again, these systems either wont be competing with AMD and Intel, or they're going to end up competing long term but periodically, so you have to factor them out evenly.

You can't just dump new technology on a market and expect high adoption rates. If Intel did this, it would probably just drop processor prices alot in dual core, and people would go for the cheap systems that would be more readily available from AMD (speaking in relative terms to today), as AMD would probably just lower prices on dual core and up quad core shipments a little bit.

You also have ramp rates and manufacturing anomalies to contend with. Intel wouldn't be able to ramp quad core that fast and ensure a cost effective and safe manufacturing process, leading to the same problem they had with the 1.13 ghz p3 (not saying it will be specifically heat related).

Overall, it would come down to if Intel felt safe lowering ASP and margins that much while increasing the likelihood for recalls, manufacturing failures, or inferior (in terms of their own line) product.

Anonymous said...

Thanks...

I think the whole reason I had started thinking about it was Servers and High end desktop is where Intel makes a good portion of its money (right?), and with all the manufacturing capacity Intel has or will have I would not be shocked if Intel was to try and push the Sever and high end Desktop to quad core as soon as possible, just like they did dual core.

AMD simply does not have the manufacturing capacity to make all server and highend desktop chips quad core, would you agree?

Sorry if it sound like the same question, its not my intention.

Scientia from AMDZone said...

That was not my understanding. The information that I saw suggested that VIA's increase was due to dumping C3 stock that was EOL.

In May 2006 Intel's cross licensing agreement with VIA expired and was not renewed, which was the reason for the forced termination of C3 shipments on March 31st 2006, as VIA lost rights to the 370 socket.

C3 was not competitive with anything from Intel or AMD as the FSB is only clocked at 133Mhz.

VIA's new chip is the C7. It uses a much faster 400-800Mhz FSB and supports both SSE2 and SSE3. I don't think this chip is in high production yet but it appears that it could start competing with lower Intel and AMD units. The problem with trying to judge its market share is that this chip is still sold in embedded as well as desktop and mobile versions. The desktop version would only be competitive with the low end Sempron/Celeron range. However, the mobile vesion at 20 watts might start cutting into the mobile Celeron/Sempron market which is higher value than the desktop low end.

The die size for this chip is very small on 90nm. It only has 256KB's total cache. Many sources list the die size at either 30 or 31 square mm's.

Scientia from AMDZone said...

Apparently someone over at Sharikou's blog was suggesting yesterday that the people over there should come over here. This may be the reason why the number of visitors yesterday jumped to 146 when it is usually around 50 per day. Apparently, they look but don't comment.

The main difference I see between my blog and Sharikou, Core2Dude, and Sharikou180 is that they seem to pick out news that supports their preferences and that's it. Their blogs consist mostly of news splashes that they think means that Intel or AMD is doing better or worse. I don't see the point of that. I prefer to do more in-depth articles (which I know is unusual for a blog). And, my biggest problem seems to be keeping the size of my articles down. My last article was indeed 2,000 words.

It doesn't matter to me if Red is a native English speaker or not. The only issue is clarity. If clarity is a problem then just writing in a simpler style would help. For example, it took me a long time to understand that Red didn't understand the memory bandwidth calculations. I just wasn't getting that from his posts. When people disagree on blog comments they will often post as rhetoric and this can be easily confused with real questions. When I understood what Red wanted I posted the calculations in detail. That's all I'm talking about. If Red finds that I'm misunderstanding his comments then he should just try again in a simpler style and maybe I'll get it the second time.

Anonymous said...

I believe I said I didn't understand how you got to your FSB numbers from my first post in the other entry, but unfortunately, you just looked for some excuse to delete it. Had you not delete it, we would know what ashenman was replying to and let the whole world know what a scathing person I am;)

I do not know how you got to your Tigerton numbers
22/10/06 5:43 PM

Again, I don't know how you came up with your Tigerton FSB numbers..
23/10/06 2:45 PM


Yada yada from me, yada yada from you til I finally get a concise answer. How more clear can I get about not understanding the FSB number?

If being a native speaker does not matter, why did you bring it up in the first place? If my English is advanced enough for rhetoric, surely I would be able to understand what I'm saying and what you're saying? I clearly told you to explain the FSB numbers, which you did not until ashenman did.. And how am I incapable of comprehending because the only thing you could point out from my post was Intel scaling back stock buyback?

Your articles are also very lengthy and convoluted and I don't see how you can not at least figure that even the most brightest of people will have trouble digesting all of it.
The final assumption..
Followed by 9 more paragraphs. I don't think you need 50 supporting details to support 1 point[that you believe "everyone" is overhyping Intel's Q4].

I'm sorry you do not like the word hot. It is in the dictionary in the context that I use it. That is not the point, if you understand it..What is there to nitpick for?

http://www.tomshardware.com/2002/06/05/via/index.html
http://techreport.com/reviews/2002q1/via-c3/index.x?pg=1
The C3 makes budget Celerons of yesteryear look like Extreme Editions.

http://www.linuxdevices.com/news/NS3116742043.html
http://www.dqchannels.com/content/reselleralert/106103105.asp
They do not specify the clock of the Pentium M..But expect 55% better on encoding, 40% on 'business', and 10% on 3D on the C7 compared to C3. And for similarly speced notebooks which would likely have greater performance, it only has 20% better battery life. And they would be competing with the rumored $450 Dell laptops.

Scientia from AMDZone said...

How more clear can I get about not understanding the FSB number?

Well, instead of saying:

"I do not know how you got to your Tigerton numbers"

You could have said:

"I don't understand your Tigerton bandwidth calculations; could you post them in more detail?"

Red, I could give examples even in this thread where I explain something and you ignore it. However, I don't want this to turn into the literary critique of Red's posts thread. I've given you suggestions which you can use or ignore.

Your links to C3 are interesting but I have no idea if you are agreeing or disagreeing with what I said or what Pointer said.

Basically. I disagree with Pointer and think that VIA has not overlapped with AMD and Intel X86 sales until now but they might in the future on low end mobile. I think the numbers are more accurate without VIA. I do agree with your VIA die size statements as beyond Wikipedia I've seen the same die size mentioned by many other places.

ashenman said...

Okay, so I need to work on my sources so I don't get crap like big dies from via (he's actually just a friend and not a source).

I think this rhetoric problem needs to die now out of apathy, as no one is going to come to a worthwhile conclusion on what needs to happen.

C7s are still leagues behind pentium Ms, and while their battery performance is not much greater than that of Intel's, it's only because their systems are normally paired with 6 cell batteries, much like many of AMD's systems.

Scientia, I agree that that's the main difference in your BLOGS, but, again, I was commenting on your replies to our comments.

Sorry for leaving you out enumae, and anyone else who ends up posting here, but I was referring to before this section became more active. I guess my point in general was, there aren't new faces, and that makes discussion somewhat less enjoyable. Thinking over the whole thing again, I guess this is the best way to do it, but I wish there was another one that were available (I've been looking over Blogger's options).

Scientia from AMDZone said...

Is there a way to have anonymous comments require acceptance on the bloggers end while letting non-anonymous ones through.

No. I can either allow all comments or registered only. I admit it would be nice if anonymous posts could be moderated while registered posts went through.

So, if we don't allow anonymous posting and more viewers don't register to post then I guess it will stay about the same as it is.

Anonymous said...

I also said in my first post on other entry 'I don't know how you got those numbers[not saying they're wrong, but the method you used]. Did you think I just said it to imply that I thought you were crazy?

I disagree that C7 will even make a dent in AMD/Intel's lower end. For all the 12W hype..Just 20% better battery life? Until they gain an obvious advantage, I don't see why anyone would bother.

I don't remeber any of your other criticisms besides scaling back the stock buyback plan that they initiated before the restructuring and my Turion comment[I would've thought that you would figure I meant price was its advantage]. Please remind me:) If you didn't feel like making a critique of my posts, why did you bother writing that? Reminds me of when you said there were a couple of reasons why AMD did not forecast numbers..And did not really say anything;)

Scientia from AMDZone said...

Okay then. Back to the topic. Any idea why Intel didn't gain either volume or revenue share? It would seem logical that if AMD were capacity limited then Intel should have picked up something.

The only thing I can figure is that this is more of an economy thing than a processor issue. In other words, if the demand for computers in general is down then the increase in volume could just be due to lower prices for no net gain.

BTW, hello to any visitors from Investor Village.

Anonymous said...

Because with Dell and other strategic partnerships they will continue to get 'artificial' spurts:) We've only started to hear of this shortage late Q3/early Q4 and if Dell is still sucking all they make, then we won't see any shifts til Q107 or so. We should be able to see AMD and Intel competing solely on technology when AMD can serve as much as Intel(when would this be btw?).

The only thing I can figure is that this is more of an economy thing than a processor issue. In other words, if the demand for computers in general is down then the increase in volume could just be due to lower prices for no net gain.
..Yes Intel/AMD are selling more at lower prices..?

Erlindo said...

We should be able to see AMD and Intel competing solely on technology when AMD can serve as much as Intel(when would this be btw?).

Once Fab-38 is fully operational. ;)

Erlindo said...

Scientia:

I'd like to see your thoughts on AMD's SOI process vs Intel's strained bulk Cmos.

I've seen some coments from intel-fanboy jumping Jack stating that AMD's SOI isn't better than intels.

How true is this?

Also, can you make an extensive overview of this in another topic?

Thanks. ;)

Scientia from AMDZone said...

Assuming that the rumors about a big leap for Intel in late 2007 are not true then it looks like AMD and Intel will hit rough parity in Q3 07. If this happens and if AMD's share projections are correct then we should see a convergence from that point. In other words, we would see steady progress by AMD from Q3 07 up to Q4 08 with AMD's reaching its target at the end of the year.

The target for AMD should be where Intel's monopoly would break. This should mean that AMD and Intel would emerge in 2009 much more alike than today. Intel would be less than 2 1/2 times AMD's size. If things actually move in this direction I would expect Intel to announce another restructuring plan in 2008 and probably big plans for CSI in 2009.

Scientia from AMDZone said...

I don't know that there is much of anything to say about Jumping Jack's statements. He is arrogant and abrasive but what he says is correct. Going by the public information presented at last year's IEEE convention the Intel tech is better than IBM's.

This may or may not be correct. Assuming that Intel was ahead in 65nm technology then their presentation should have been further ahead. Is it now? We don't know. Also, he suggests that when the next convention occurs we can get another update. Again, this may or may not be the case. If Intel, IBM, and AMD only present papers on 45nm tech then we won't know how their 65nm compares. We should get some idea about the rumors for Intel's 45nm tech though.

Also, some of the techniques begin to bend when you get small enough for quantum effects to kick in. If the rumors about Intel's 45nm improvements are correct they would have to be due to quantum effects.

I do know that the great majority of companies have abandoned the high K path that Intel is pursuing. However, it is possible that Intel has found a suitable high K material with a good bandgap (as good as SOI).

The papers are a good starting point. In the 2005 papers, Intel was definitely ahead. AMD claims that they have continuous improvement of transistors. So, we don't know if the performance gap narrows over time. In other words, the papers are leading tech but I don't know of any public comparison of ending tech like Intel's current 90nm versus AMD's 90nm.

TheKhalif said...

As usual you are not afraid to tell the truth. I like it. Most peopel will blindly believe the Intel hype about AMD. As you said, AMD breaking into China through Lenovo and the other OEM(very chinese name) will ensure that their volume doesn't go down even if it doesnt go up much.

300mm 65nm wafers will make more than twice the amount of chips as Fab30 or Charetered.

AMD made some really smart moves like cancelling 1MB chips and all of 939.

4x4 is a great idea and I believe it will catch on in the mainstream before too long.

People who want the 2 proc wksta will love that they can save $1000 and still get 2 procs.


Lots of folks complain that 4x4 was a knee jerk to Core 2 but how else could they get a NOTICEABLE JUMP in FX perf? It was either get an extra 1GHz or add a socket.

The ATi deal will also do a lot to ensure that Vista loves AMD. And with nVidia basically now being independent(AMD's purchase doesn't change their share + or -)

They will be in the position to get Torrenza diectly from AMDs ATi division.

Their cooperative business model will cost Intel BILLIONS. I don't think they will Ch.11, but they will be bleeding money like a hemophiliac.

They have yet to realize that there is now a duopoly and MAD can't be pushed around as before.
Opteron made sure of that. Not to mention Dell.

And all this talk of shortages, it came out that they were complaining about 939 which has been cancelled. Newegg has retail 5000+ for $399 and OEM 5200+ for the same price.

And the beauty of AMDs platform is that 939 can be made into AM2 with the change of a socket, so mobo manuf can use the same board layout for all desktop mobos.

Anyone who is honest has to admit that AMD is not going anywhere but up. It is always that way. For example, Windows will NEVER gain share but MacOS can.

Just like IE and FF. FF has yet to lose any share while IE has only been losing share.

I have long related Intel\AMD to majority\minority in a society.
The majority can do no wrong but the minority can do no right.


ALL HAIL THE DUOPOLY!!!

hyc said...

thekhalif said:
And the beauty of AMDs platform is that 939 can be made into AM2 with the change of a socket, so mobo manuf can use the same board layout for all desktop mobos.

Changing S939 to AM2 also requires this tiny little detail of routing 240 DDR2 DIMM pins instead of just 184 DDR, but hey, no big deal right?

TheKhalif said...

Changing S939 to AM2 also requires this tiny little detail of routing 240 DDR2 DIMM pins instead of just 184 DDR, but hey, no big deal right?

It's still the same mobo basically. That was the point. if you start with the board before RAM traces are added you just do what you would do for AM2.

Of course you can't just rip the socket off, but I would say that it would be much easier to change over prod to AM2 from 939 than 975X from 945.

ashenman said...

Just more proof of Scientia's lack of bias. He gets complaints from both AMD and Intel fans.

Buy yay, we have a new commentor or two, so I'm not complaining.

I'd like to ask what of Intel's lies we've believed? I'd be sad if Intel were in a more precarious situation than it is, as an unbalanced market is my main complaint against Intel, and it would be no different if AMD took Intel's place.

Dr. Yield, PhD, MBA said...

Scientia, you have a fundamental flaw in your buyback logic. It isn't the $ spent on share buyback that matters to the street, but rather the percentage of diluted shares retired. Remember, price of a share is primarily a function of "value" or free cash generated on a per share basis. By removing shares from the market (+dilution), you increase the value of a share. So let's look at the numbers:

Diluted shares outstanding (millions):
Q4 2003: 6671
Q4 2004: 6352
Q4 2005: 6081
Q3 2006: 5832

Good trend, but does it mean anything?

Shares removed from circulation:
03-04: 319
04-05: 271
05-q3 06: 249

Ok, this number is trending down, although to equal last year's numbers at $21/share, Intel only needs to spend another $462M. To meet 2004 reduction would cost $1.47B.

Percent of shares removed from circulation:
03-04: 4.8
04-05: 4.3
05-q3 06: 4.1

Again, a downwards trend. But, to achieve the 4.8% reduction of 2003-04, Intel would need to buy 31M shares, which at $21/share is 651M- far less than the 1.47B above, and way less than the 3.1B to bridge the gap with 2004 spending. 2005 spending? To achieve the same effect for shareholders as 2005, Intel needs to retire 12.2M shares in Q4, for about $255M, less than the $462M above, and waaaaay less than the 6.2B deficit you imply.

Conclusion? This year's reduced share prices have enabled Intel to make sure that shareholders are not losing out on the buybacks this year and spend less cash in the process.

ashenman said...

By saying "that matters on the street" are you saying that that's what matters to investers? Cause we're talking about the companie's health, not stock health.

I'm not quite into trading and the financials involved with shares, but if Intel buys back the shares that doesn't necessarily mean that they're out of circulation does it?

Dr. Yield, PhD, MBA said...

Original posting:This would be a 50% reduction from the 2005 levels and something the stock market surely would not overlook. This could easily cause Intel's stock to slump when the end of year results are released(which will be sharply down from last year).

Ashenman queried:By saying "that matters on the street" are you saying that that's what matters to investers? Cause we're talking about the companie's health, not stock health.

Ashenman, I was responding to the above claim. The simple fact of the matter is that Intel's stock will not (or should not) tank because of a reduced buyback in $ terms. Because in % of shares removed from the market, Intel has been relatively consistent. By reducing the number of shares, the company has increased the value of a share- more revenue, more income, and more equity per share held.

And yes, if Intel buys those shares back, they are either retired or held as "treasury stock". Either way, they are no longer part of the outstanding or diluted share totals. The company can always reissue them, but at the expense of dilution of shareholder value. This is true whether they use them for cash, employee comp (options), or acquisitions (a la AMD buying ATI by issuing new shares)- whatever the purpose, they do so at the expense of dilution.

Scientia from AMDZone said...

Intel was up in both 2004 and 2005 with year over year growth. This will be the first time since 2001 that Intel has shrunk.

I don't have any confidence that Intel can affect its stock price by spending less than $500 Million. $1.4 Billion would be more like it. BTW, if stock value is inversely proportionate to the percentage volume of dilluted shares in circulation then shouldn't Intel's stock value be higher now?

My guess is that Intel's stock will take a hit when they release the worse year end report in five years. I wonder if the market will see AMD the same way when it has doubled in size in two years. I guess we'll see what happens.

Dr. Yield, PhD, MBA said...

BTW, if stock value is inversely proportionate to the percentage volume of dilluted shares in circulation then shouldn't Intel's stock value be higher now?

Clearly that isn't the only indicator. But P/E ratios are clearly an important indicator of value. By reducing the number of shares, E increases, lowering the P/E ratio. A P/E ration of 30 might be fine for a growth company, but not for a company stagnating- that might merit a ratio of 15.

If the stock takes a hit because of the year end report, it won't be because of the buyback numbers. It will because prospects for growth merit price premiums. If the street doesn't see a growth story, expect the price to drop. See DCF (discounted cash flow) analysis for more treatment of the subject.

Anonymous said...

So while AMD/INTC are at current values, people are suddenly going to dump/pump because of end of year reports?..Unless you're saying that you believe AMD will have better than expected Q4 numbers and Intel to have an abysmal Q4. Last time AMD beat estimates, they got tanked on margins anyhow.

ashenman said...

It depends on how intelligent investors are in general. Scientia is just assuming they aren't, which may or may not be a good assumption (though it's generally a good one). If investors have just been looking at QoQ improvements expecting a big increase or at least respectable increase in q4, they'll obviously be disappointed, and sell.

Analysts are giving Intel a nice Q4 last I heard, so that's why people will have this expectation.

R2K said...

: )

Scientia from AMDZone said...

Well, looking at the rankings:

Top 15 Semiconductor Rankings

AMD has moved up 3 places from 16th to 13th. With an average growth of 8% in the top 15 AMD's growth was 35% while Intel's was -11%.

-11% growth doesn't suggest a lot growth potential to me while 35% suggests quite a bit. In terms of real growth potential clearly AMD is well ahead of Intel.

Howevever, people who buy stocks don't go by growth potential of the company; they go by growth potential of the stock. AMD's rapid growth won't slow until it hits the 3rd ranked position however its stock value may not change that much due to the fact that it puts so much of its income into capital investments. This won't change until AMD builds the NY FAB.

pointer said...

Assuming that the rumors about a big leap for Intel in late 2007 are not true

Just curious, what exactly the rumor is? what kinda big leap? care to share the details?

Scientia from AMDZone said...

October 11, cpu roadmap

They are suggesting the 45nm version of Conroe will draw 57 watts and clock to 4.0Ghz.

October 20, cpu roadmap

Here they have the same 4.0Ghz speed for 45nm dual core. However, they have increased the estimated speed for the 45nm version of Kentsfield to 3.73Ghz. They show the dual core as having 6MBs of cache and the quad with 2x6MBs.

Azary Omega said...

Scientia from AMDZone said...

October 11, cpu roadmap

They are suggesting the 45nm version of Conroe will draw 57 watts and clock to 4.0Ghz


I call a bull on this! No for real people:
6MB cache x 4.0 GHz x 45nm = 57W
while
4MB cache x 2.66 GHz x 65nm = 65W

I'm quite sure move to 45nm wont give that much improvement. Unless intel will use some new technology in their process.
Do they?

Anonymous said...

We know about the rankings now. Suddenly they're going to pay attention to it at the end of the year? We've also had year over year comparisons with each earnings release and 3/4 of their end of year report is here. What is so special about the end of year report?

http://www.gamepc.com/labs/view_content.asp?id=c2dm&page=4
There's Merom, for significantly less wattage, is on par with Conroe. With T7600, 2.33, 34W, a 71% increase to 4.0 would make for a 58W TDP. Very theoretical, but there it is:)

http://www.tomshardware.com/2006/09/25/green_machine/page4.html
The 89W X2 3800+ consumes less than the 65W X2 4600+ EE anyhow. I call all TDP ratings BS; mobile, desktop, high end should be fine enough:)

http://www.dailytech.com/article.aspx?newsid=2649
With P1266, Intel shifts away from Silicon Dioxide gate dielectrics -- a process the company has used since the mid-90s -- to High-k dielectrics. With new dielectric techniques, the company will also revamp its gate electrodes to metal instead of Polysilicon derivatives. The last major materials change of such magnitude occurred when Intel moved from Silicon to Strained Silicon in 2002, which is still slated for use in P1266 and beyond.

http://images.tomshardware.com/2006/11/03/amd_mainstream_dualcore.jpg
BTW, I'm curious why AMD has Kuma for 2.0-2.9 yet is still introducing 90 nm 89W 6000/5800 3.0, 35W 4000/4400 2.1-2.3 in Q307.

ashenman said...

AMD typically ramps new process performance slowly to ensure they get more yield out of it. I'm sure many of those chips will overclock like crazy as well as undervolting like crazy. Like my 3800x2 that is overclocked 500 mhz while undervolted .125 volts (I hope this explains why the 3800 89w part may not have needed to use all 89 watts of power). AMD will get more performance out of 90 nm as well, because of how they add enhancements dynamically to their processes. They wont be able to implement all the ones they had on 90 nm to 65 nm right away, so it wont see immediate benefit.

Intel is looking into high k gates, which means they'll possibly have less leakage across their transistor's dielectric compound (the chemical that makes the transistor stop current flow when it doesn't have an on signal). This could be both good and or bad. If it works at the testing level, but on broader application ends up like their 90 nm process and leaks voltage like crazy, then they're stuck with a poor design with no back up subsystem (like SOI or SSOI) to mitigate this disadvantage. If it works at both levels then they have an awesome processor and I see no reason they couldn't almost (because those numbers are still way out of reach if this did happen) get those results. If it doesn't work at the testing level, then they may or may not have voltage leaks/a really hot processor but definitely wont make those numbers. Even with all of that considered, there are numerous other considerations like manufacturing ramp times and the like, which means that it wont happen at a significant volume. The same exact thing could happen with the metal gate electrodes.

All in all, Intel is just trying to paint a bright picture for distributors so they can continue to make long term contracts that will concern these or current product lines and to inflate stock so that buyback can be minimized.

pointer said...

October 11, cpu roadmap
...
October 20, cpu roadmap
....


Thanks.

Azary Omega said...
I call a bull on this! No for real people:
6MB cache x 4.0 GHz x 45nm = 57W
while
4MB cache x 2.66 GHz x 65nm = 65W

I'm quite sure move to 45nm wont give that much improvement. Unless intel will use some new technology in their process.
Do they?


While i'm not sure if the rumor is 100% true, but it is possible. Intel has a 'new' transistor for the 45nm version while AMD use back the 90nm transistor in their 65nm. The previous statement given was purely from my head, read some where but forgot where. Anyone comes across the material or can prove what i said is wrong, please do so.

definition of 'old' - if you cut it and look thru the electron microscope, the structure and material looks similar to the previous generation. it is a some sort of shrink. Again, i can be wrong due to my potential faulty brain memory :)

Scientia from AMDZone said...

The reason the year end results are important is because Intel could still have a good 4th quarter or a more disappointing quarter. There was a lot of slack in their projection.

AMD is also moving to metal gate connectors so this isn't much of a difference.

The gate material is a fundamental problem. Intel would like to make the gate material thicker to stop leakage but then it would get clobbered on capacitance. To avoid higher capacitance they switch to high K however most high K's have a lower bandgap than SiO2. Apparently, Intel is confident that they've found a high bandgap/high K material. AMD will use a more pure SiO2 to reduce leakage. The really novel stuff hits later when they try winged or tri-gates.

Anonymous said...

Maybe this helps, I do not completely understand all of it, but it does explain what you guys are talking about...

-[LINK]-

ashenman said...

I think the paper you linked shows Intel's fundamental development problem, enumae. While it's nice to develop completely new technologies, doing that every time you find a hitch in one of your current technologies instead of resolving or utilizing that hitch is expensive and often leads to differences in what you can produce in your lab and what you can produce in volume.

Intel has been able to execute this strategy very well previously, because of how much capital it had. It is still able to do this, but I think it needs to develop some of AMD's more cost effective tactics if it wants to adjust better to its current market standing.

Anonymous said...

Ashenman said...

"I think it needs to develop some of AMD's more cost effective tactics"

I am interested in what these are, could you explain?

Thanks :)

Erlindo said...

Ed is once again talking doom and bloom on AMD. He's even repeating the non-sense of AMD having to be bought by someone to become an intel-killer.

To achieve the above mentioned, all AMD needs to do is ramp as fast as possible Fab-38 and start the other fab in NY.

Even with Fab-36 and Fab-38, AMD can achieve more than 40% market share.

Sometimes I believe Ed is up to something. He should stop "smoke-it-green" for good. ;)

ED is talking BS

Anonymous said...

erlindo said...

"To achieve the above mentioned, all AMD needs to do is ramp as fast as possible Fab-38 and start the other fab in NY."

To become an Intel killer they need much more than just 3 FABS.

Think chipsets, think integrated graphics, think discrete graphics, think R&D, until they can supply (just guessing here) about 50% of the market with an integrated solution Intel will remain dominant.

Ed made some very good points about AMD's money situation.

"Even with Fab-36 and Fab-38, AMD can achieve more than 40% market share."

Actually they cant, and if they thought they could Ruiz would have said something more than "hoping to achieve 30% leaving 2007", something along those lines, but thats what AMD is hoping for.

Also keep in mind Core 2 Duo is ramping (slowly) but the end of Netburst will be huge benefit to everyone, even me, still got one :)

Erlindo said...

Actually they cant, and if they thought they could Ruiz would have said something more than "hoping to achieve 30% leaving 2007"

Sure they can ;)

If they (AMD) achieved 18% market share with just one 200mm fab, imagine what two 300mm fabs can do at full ramp!

ashenman said...

By cost effective techniques, I mean developing technologies that compliment their process.

Dumping a process every time you hit the limit of running it naked (meaning without improvements) is expensive, because you have to buy new toolsets, and change out packaging systems. Especially the way Intel does it, where when a fab's process becomes obsolete they either make chipsets with it or quit using it, meaning they have to build new fabs all the time.

Ruiz is assuming they wont have full capacities from both fabs because they'll be updating fab 30, while building fab 38. It'll be very interesting to see where it leads.

While I think the core is an awesome processor, I really couldn't bear to see Intel take back full market ownership like they had during the netburst and pre-p3 era. A lack of competition really stifled innovation, and lead to a lot of dealmaking meant to benefit companies without benefiting consumers. Does this mean I think Intel is evil? No, they're doing what any company in their position would do, but it's not healthy or sustainable.

Anonymous said...

Erlindo said...

"Sure they can ;)"

Assuming on a 200mm wafer at perfect yields the get 115 X2's, Opterons, etc...

Assuming they do 30,000 WSPM that would equate to about 3,450,000 processors at 90nm for FAB 30.

Assuming on a 300mm wafer at perfect yields they get 274 X2's, Opterons, etc...

Assuming they do 15,000 WSPM and make the transition to 65nm with half of there wafer starts that would equate to about 2,055,000 processors at 90nm and 2,887,500 processors at 65nm for FAB 36.

Totaling 8,392,500 X2's, Opterons, etc... each month. Totaling 25,177,500 processors per quarter, or about 100,710,000 processors per year.

Projected number of processors made for 2007 is around 250 million, so that 100,710,000 processors would be about 40% of the market share at perfect yields.

Keep in mind that FAB 30 will go to about 50% capacity while converting to 65nm and 300mm wafers, and FAB 38 is not expected to have 65nm crossover until around June, at which time the conversion of FAB 30 to FAB 38 begins.

Now factor in yeilds, I have no idea, and neither does anyone who does not work for AMD, but we wil guess 70%, from what I have read some think that is good and some think its more like 80-90%, new node and misc factors well go with 70%.

100,710,000 * 0.70 = 70,497,000 processors or about 28% of the market share, I understand these are all dual cores, and they are all just estimates, but this would put them about where Ruiz said they would like to be.

So no 40% is not really achieveable, not with what I am looking at.

If some one sees an error please tell me, I hate having missed something.

Thanks guys.

PS: We haven't even factored in quad cores, which is where I think Intel will be able to really show the manufacturing muscle, better processor or not :)

Also if you guys want some links to the basis for my numbers just let me know.

Azary Omega said...

Yes Enumae, you made a mistake in that all 90nm dual cores had (have) 2x1MB of L2 cache and revF (65nm shrink) has 2x512 of L2

Ho Ho said...

"all 90nm dual cores had (have) 2x1MB of L2 cache"

From what I've seen it is quite difficult to get 2x1M L2 x2 nowadays. Most have 2x0.5M.

Anonymous said...

Azary Omega said...

"Yes Enumae, you made a mistake in that all 90nm dual cores had (have) 2x1MB of L2 cache and revF (65nm shrink) has 2x512 of L2"

I will re work the numbers tonight.

Thanks.

Ashenman I did not forget about you, thanks for explaining, but I will get into details later as well.

Scientia from AMDZone said...

AMD did produce 45 Million in 2005 with just FAB 30. AMD should make 75-80 Million processors in 2007. I'm not sure the 250 Million number is correct as this would be almost no increase from 2006. I think AMD will only be at about 27-28% of the volume by end of 2007. By end of 2008 it could be slightly over 30%. Part of this depends on how fast the market expands. If the market expands faster AMD will have less; if it expands slower AMD will have more.

Anonymous said...

Here is where I got the basis for my 250 Million.

-[Link]-

If you have something newer I would love to see it, I do not like to assume.

Thanks

PS:If the market expands faster AMD will have less; if it expands slower AMD will have more.

Completely agree :)

Anonymous said...

Numbers revised... Die size has gone from 230mm2 (1MB cache) to 199mm2 (512KB cache)

Sorry if bits and pieces of text are out of place, in a hurry :)

Assuming on a 200mm wafer at perfect yields the get 133 X2's, Opterons, etc...

Assuming they do 30,000 WSPM that would equate to about 3,990,000 processors at 90nm for FAB 30.

Assuming on a 300mm wafer at perfect yields they get 316 X2's, Opterons, etc...

Assuming they do 7,500 WSPM at 90nm 2,370,000 and 7,500 WSPM at 65nm 3,345,000 processors FAB 36.

Totaling 9,705,000 X2's, Opterons, etc... each month. Totaling 29,115,000 processors per quarter, or about 116,460,000 processors per year.

Projected number of processors made for 2007 is around 250 million, so that 116,460,000 processors would be about 47% of the market share at perfect yields.

Keep in mind that FAB 30 will go to about 50% capacity while converting to 65nm and 300mm wafers, and FAB 36 is not expected to have 65nm crossover until around June, at which time the conversion of FAB 30 to FAB 38 begins.

Now factor in yeilds, I have no idea, and neither does anyone who does not work for AMD, but we wil guess 70%, from what I have read some think that is good and some think its more like 80-90%, new node and misc factors well go with 70%.

116,460,000 * 0.70 = 81,522,000 processors or about 32% of the market share, I understand these are all dual cores and all with 512KB caches, but they show a good idea of what AMD is going for, but they will still be short of the 40% mark that Erlindo was talking about.

Again if some one sees an error please tell me, I hate having missed something.

Thanks guys.

------------------------------------

Ashenman said...

"Ruiz is assuming they wont have full capacities from both fabs because they'll be updating fab 30, while building fab 38. It'll be very interesting to see where it leads."

From what I understand they are not building FAB 38, that is the name that FAB 30 will be given when the conversion is done (65nm and 300mm wafers).

Scientia from AMDZone said...

That link is a year old. But did you notice this line:

With just a few days left in 2005, IDC is upping its forecast for 2005 growth

When IDC ups its forecast for 2005 on December 22, 2005 this doesn't exactly give me confidence in their 2007 forecast.

Scientia from AMDZone said...

Well, in your estimate you didn't scale down FAB 30 and you didn't allow for the differential from FAB 36 while both ramping and converting. Also, I don't think you allowed for either opteron or quad core. So, again, I'd say my estimate of 75-80 Million is closer. AMD's own estimate is that they will do 100 Million in 2008.

ashenman said...

That would be pretty awesome if they could pull 100 million. In a recent interview Hector said that fab 30 would drop production a quarter while changing production. I think there was a bit of miscommunication, as Hector was apparently quite pleased with this report while the reporter thought he was insane. Point being, I think they're only dropping fab 30 production to 75% as they convert, and that they'll probably end up converting very quickly. Either that, or they'll drop to 25% production, and end up converting really fast, and we'll have some dip in the middle of 07, and then an explosion of AMD parts in time for the holidays.

Scientia from AMDZone said...

Hector said that fab 30 would drop production a quarter while changing production.

No. Production in FAB 30 begins decreasing mid Q2 07 and doesn't increase again until Q1 08.

Point being, I think they're only dropping fab 30 production to 75% as they convert

No. Production in FAB 30 will drop to 40%.

, and that they'll probably end up converting very quickly. Either that, or they'll drop to 25% production, and end up converting really fast, and we'll have some dip in the middle of 07, and then an explosion of AMD parts in time for the holidays.

None of the above. Production in FAB 30 will drop sharply from 100% to 40% by Q1 08. During this time production in FAB 36 will continue to increase from 15K to 20K wspm.

FAB 30 will drop by 18K wspm.
FAB 36 will increase by 5K. However FAB 36's 5K is equal to about 20K of FAB 30's. This just slightly overcomes the loss from FAB 30. However, AMD is currently building an expansion area which will be ready about the time that FAB 30 begins scaling down. This expansion area will give AMD the extra capacity it needs to keep growing as FAB 30 drops. This area is equal to about 5K wspm. Perhaps this is the "shuffle" that Ed (Overclockers) was referring to.

ashenman said...

I'm sorry, but straight from the horses mouth they're dropping 1 quarter production. Either to, or by, because we don't have a very well explained answer. So I want to know where you get your 40% figure, because if it's historical, then you fail to notice the fact that the situation we're currently facing is unprecedented for AMD, and if you have a link where they say otherwise, I'll have to dig up mine and we can compare dates, and if it's just speculation, then we don't need to speculate to that degree, because we have something resembling an answer.

ashenman said...

By the way, by one quarter, I'm directly quoting Hector, and I mean 25%, not a financial quarter.

Anonymous said...

*cricket*
...
Someone: Frankly feeding ATI Gpu & Chipset with an AMD 25w Turion single core part (MT32, MT34, MT37, or MT40) to Apple for low end macbooks for the $700 to $1200 range makes sense.

Since you agree with this person, I'll address it as if you said it since it's your blog.

First, Hector saying.."Why would they want to be held hostage like everyone else has been?" Sorry, but calling a possible customer, especially the pretentious Apple of all, hostage does not seem like anything is in the works soon.

http://www.gizmodo.com/gadgets/press/apple-iphone-order-for-12-million-confirmed-215105.php
http://www.reghardware.co.uk/2006/11/16/apple_amd_laptop_claim/
A Taiwanese source talking about Japanese suppliers.. Seems more like Hon Hai ramping up for iPhone.

What advantage would an ATI platform have over an Intel one? Better graphics? Because we all know of the great supply of games on Apple:/

Laptops in the $700-$1200 range? $700/single core would further make Apple as 'just another PC company'. Considering that Apple achieved 12% share recently, I don't know why they would bother. Also, Apple does not like presenting too much choices for the consumer, especially ones that would overlap.

You: Agreed. The extra SSE power of Merom is not helpful at the moment because Merom is unable to match the low power draw of Yonah. It appears that Intel will be unable to replace Yonah until Merom transitions to 45nm. As an indispensible part of the Intel lineup it appears that Yonah was not such a yoni afterall. However, Turion is competitive with Yonah.

A general usified mobile core would obviously not be as efficient as an optimized mobile core. At the mainstream level(non low voltage), the difference is negligible unless you'd like to prove otherwise.

http://www.tgdaily.com/2006/11/07/intel_mobile_processor_roadmap_update/
http://www.tgdaily.com/picturegalleries/gallery-200611131-11.html
http://www.theinquirer.net/default.aspx?article=35389
Core Duo is currently 30%, Core 2 Duo[mobile] is 40%. In Q107 Core 2 Duo will be 60% and Core Duo will be 15%. Core 2 Duo will be 80% by Q307. Pentium D and Yonah are to die around mid07. Does that make Pentium D indespensable because it'll stay in production for a while:/.

ashenman said...

While I think the power arguments are a bit more disputable, the graphics argument is important, as Intel graphics aren't powerful enough for serious photoshop work (should I even both explaining why that would be important to a mac user?), they already buy alot of ATI graphics, and don't want to do Nvidia for obvious reasons (which I bet Intel will try to push them to). Plus, if they buy on an AMD/ATI platform, It'll probably be cheaper than if they try to mix ATI with Intel.

Mac would still have a better image than most pc companies because their cases look much better and they still have their own operating system.

As for Hector's comment, I can't believe people still act like these companies are little children that get pissed off at each other. The only reason Apple might be concerned about what AMD said is because they don't want to send mixed signals to their customers, and last I remember, Mac users weren't big hardware enthusiasts, and I'm sure very few read the Inquirer. Even then, this may be mac answering what Hector said by trying to show what customers are actually interested in hardware that they are willing to give them a choice (highly unlikely for obvious reasons). Mac is expanding its audience rapidly by lowering prices and increasing performance. By lowering prices even more without offering a severe decrease in performance, they gain even more market.

Anonymous said...

I've looked up GMA950 Photoshop numbers and they are indeed abysmal. But surely any one that is in serious Photoshop work would do their fancy stuff on desktop/workstation or at least a MacBook Pro which don't have IGP?

http://www.macobserver.com/article/2006/07/19.16.shtml
With the terrible GMA950, they achieved 12% notebook retail share recently. Am I the only one that thinks 12% is huge btw? Not all of us are Photoshoppers:)

http://www.apple.com/macpro/graphics.html
http://www.apple.com/imac/graphics.html
http://www.apple.com/macbookpro/graphics.html
They seem to have a nice mix of Nvidia and ATI, whatever's appropriate for the situation.

True, they have the used to be almost cheap though not really since HPs are as low as $249 Mac Mini and they're still considered 'fashionable'. Though I personally think that Mac Pros are ugly. And iMac G4 > iMac G5:D With HP at $399 notebooks, you'd think Apple would need to compete more but they seem content with that 12% share with an entry level $1099, not that I wouldn't want that to go lower.

I'm not saying that AMD would piss off Apple with that comment [though with the proud way they talked about x86, other PCs, Windows in the past, I wouldn't be surprised], but the aggressive way that he suggested something that should be kept quiet does not seem like anything is in the works soon.

Intel CPU+IGP vs DAAMIT CPU+IGP? I would think that Intel would be cheaper because of manufacturing, even if they charge more[which they can dump to Apple]. Though significantly better than Intel's IGP, I don't think that's enough for them to fiddle with their product line, unless they replaced all MacBooks[not Pro] with DAAMIT [wouldn't want cheapo DAAMITs being graphically superior than other Intels].

http://www.macrumors.com/pages/2003/09/20030917033706.shtml
So in summarization, I don't believe Digitimes that Apple is going AMD because ?a certain MLCC? is seen on the rise while iPhone is ramping, Hector saying silly stuff, AMD already juggling with everyone even more with Dell, no compelling reason to go DAAMIT. Oh yeah, link above shows why we should remain skeptical of Digitimes.

ashenman said...

I guess my point is, for those that want serious photoshop work, want the status symbol of a mac (because ya, they aren't the "best" looking, but it still looks different than almost anything else out there that isn't trying to copy them) but can't afford the non-igp solutions from macintosh, then the DAAMIT solution is best.

Your point about there current market seems misplaced as it strengthens my argument that they're trying to gain new marketshare by lowering their entry level prices. To make the pros better than their entry level pieces, they only have to make those non-igp only, and voila, you solve the problem in the performance deficit. The other thing to consider is macintosh's relatively slow development cycle. I have the feeling that the whole AMD plan isn't short term, and is looking to a later product line up, when AMD theoretically could hold the performance crown, and will definitely have a much better igp solution than Intel.

"Intel CPU+IGP vs DAAMIT CPU+IGP? I would think that Intel would be cheaper because of manufacturing, even if they charge more[which they can dump to Apple]."
What does that even mean? Are you saying it would be cheaper for macintosh to have a more unified product lineup in concern to manufacturing, even if they're being charged more for all of the components? This is a marketshare gambit they'd be playing, I'd assume they wouldn't care about the extra cost if it let them get into an area of the market they haven't even touched before.

The graphics solution argument is pretty pointless since we're talking about notebooks here, and the only thing mentioned for their notebooks is the x1600 mobility. Again, for obvious reasons, they wont be going nvidia.

"but the aggressive way that he suggested something that should be kept quiet does not seem like anything is in the works soon."
Again, I don't understand what you're saying. If you're suggesting that the fact that AMD feels that Intel is monopolistic should be kept quiet, then you really need to look back to their previous statements, however, since you actually do that in your argument, this is completely counter to your own argument concerning their past statements. If anything, this hints at absolutely nothing, because either Hector is saying that to make Apple look bad later for not hooking up with AMD, or he's saying that to make Intel look bad when Apple does hook up with AMD.

Also, red, a summary should go over material you actually argued, instead of bringing up a completely new argument or mentioning ones you didn't even mention in the body of your argument. Plus, you need to make sure your arguments aren't contradictory.

Anonymous said...

They didn't get 12% share with "cheap" $1099, when in the past they were selling for $999.. They got it with brand recognition from iPod. And also possibly performance parity with Intel. More so on the iPod.

Serious photoshop work, want the status symbol of a mac, but can't afford the non-igp solutions from macintosh, then the DAAMIT solution is best.

Serious Photoshop work and can't afford non IGP doesn't go together:/

So why would DAAMIT be cheaper than Intel is what I'm asking when Intel has better manufacturing? Is this just an assumption?

Sorry, but you said for "obvious" before on ATI mobile and didn't explain. Please explain:) You make it seem as if it were political differences yet they use Nvidia on desktops.

What I'm responding to is that some believe the Digitimes report that Apple is launching AMD stuff soon to be true and use the recent Hector comments to support that.

AMD has spoken about Intel's evil monopolistic behavior in the past. There are better ways to have said this then calling a pompous potential customer as being held hostage. But that's not the point, the point is that something as huge as an Apple/AMD partnership wouldn't likely debut at some dinner.

Scientia from AMDZone said...

So I want to know where you get your 40% figure

From AMD. Previously, AMD said that production in FAB 30 would drop to 50%. Recently, they said it would drop to 40%. However, since FAB 36 will be increasing and it will be about 3X the output of FAB 30 production will still increase.

Anonymous said...

No duh it came from AMD:/ A bit late, but one of you guys should try to substantiate your claims with a link or something;)

Scientia, your latest article, AMD doesn't suck, was vastly improved compared to the other Intel does suck articles:) And you threw in some pics to not make it a total drag to read. BTW I will not try and refute anything that I don't know about anyways:)

Anonymous said...

http://theinquirer.net/default.aspx?article=24756
I have no idea what you are talking about with PCIe but maybe that will help:)

http://www.xtremesystems.org/forums/showpost.php?p=1491785&postcount=24
Anything wrong with his assesment? I thought HyperThreading was for multitasking not to mask memory constraints:/ POWER5/6, Niagara are also multithreaded.. And we know little of HT2, though I have a feeling it may be Speculative Threading/Mitosis.

BTW, may I ask that you not take one person's words and categorize them into a whole, like whoever you see as so anti AMD to categorize all Forumz members:)
Naturally though, someone suggests that any information that favors AMD is mere FUD. Woot, singling out a person rather than categorizing everyone:)
http://forumz.tomshardware.com/hardware/thoughts-AMD-quietftopic-209658-days0-orderasc-50.html
Are you talking about that btw?
Last I checked, AMD64 == EMT64, for all intents and purposes.
I agree with r00t, what does the branding of it matter if they work the same?

One of the problems with Forumz is that many people there tend to believe the testing that Toms Hardware Guide does.
Really, like how the recent server comparison was thrashed? I can think of ochungry, HelloKitty, BaronMatrix, Parrot, Serge84, MadModMike, MrsBytch, 9-inch, nn_step, many from a certain zone;) as prominent AMD fudders. Who are these silly folks that care whether or not Intel is better?
Some people there still insist that an E6300 is faster than an FX-62 based on this.
..
http://forums.anandtech.com/messageview.aspx?catid=28&threadid=1941706
I'd like to see the some that believe E6300 > FX62:)

Like Sharikou, whom none of us should bother wasting our brain on, 9-inch had a reputation of.. Stupidity and used your intellect to further his 'Intel is Suxx0rz!' agenda. You're defending a guy named after a certain anatomical part?
Really, why would anyone bother reading whatever he/you had to say with his past?

And by Core 2 being better than K8 in SSE, you mean the benchmarks and apps that matter?;)

A: K8 has better memory bandwidth and lower latency than C2D.
B: Yeah, that's true, but that doesn't matter because THG's testing proves that even Kentsfield doesn't need the memory bandwidth.


Really? I usually see B as memory benches aren't worth a crap if we don't see them in apps.

You just don't tend to see the same kind of protectiveness here for AMD or the same kind of wild optimism.
*rubs eyes*
http://amdzone.net/index.php?name=PNphpBB2&file=viewtopic&t=10276&postdays=0&postorder=asc&start=15&sid=ce81f9be94936f7881e40c00a7d65a55
http://www.hardforum.com/showthread.php?t=1120674
BTW, I don't remember breaking any rules:) Someone seemed to get upset because I called them out for their condescendence:)

Scientia, could you switch to Blogger Beta btw? I tend to start typing here before I'm logged in, then have to copy everything, then login, where as with Blogger Beta I can do both at the same time.
Happy Thanksgiving btw:)

Anonymous said...

http://www.hexus.net/content/item.php?item=7316&page=3
http://www.theinquirer.net/default.aspx?article=34433
http://www.theinquirer.net/default.aspx?article=35902
Hexus's latest article seems to confirm Charlie's that HT3 won't be AM2+..
So although the initial Quadfather may not win a shootout with Core 2 Quad hands down, its sequel, Quadfather: Part II, could be the one which really makes a wise guy out of Intel again. And there's a third part of the trilogy mooted for the end of 2007, too, with DDR3 and HyperTransport 3.


So silly Scientia, to address me while I'm anonymous on 180 yet ignore me on your own blog:/

You should be able to tell who I am if it means so much to you;)?? If 180 provides me with the convenience to not spend 5 seconds on logging in then I will:D Argue with the message not the person:/

http://scientiasblog.blogspot.com/2006/10/real-forecast.html
Maybe you should read your predictions again but I don't see anything insightful:/ It's obvious that if 1 Prescott isn't good, 2 isn't going to be much better. They've had mobile on desktop before so it's expected for Yonah to also be on the desktop.
http://www.theregister.co.uk/2005/03/15/amd_turion_bench/
http://www.theinquirer.net/default.aspx?article=14257
With the ownage of Turion by Pentium M and the wall of heat approached by Prescott, it was obvious that they weren't going to continue the Netburst path.

I stand by my case that you have nothing to brag about unless you predict something like AMD and ATI as well as correct on many other unforeseen.

Anonymous said...
Scientia mite go around proclaiming his unbias but he is at heart an amd fan boy but def better than sharikou cuz unlike the doctor atleast his theories are not all that far fetched.

Are you upset by that btw? It's not me, as I'm sure you are able to tell by now that I can spell:D I'm not the only one that can see you slant your blog towards pro AMD/anti Intel stuff.

With my advanced Googling skills, I've found these comments from Sharikou about you. Come on, and you still insist to us that you have no preference towards AMD/Intel? Even your own supporters can see.

I think Scientia is a couple of notches above sharikou in the intelligence front. He's definately an Amd fanboy like sharikou, but unlike sharikou he know how to present an opinion without comming across as a cartoon caricature, which sharikou has now become.

WTF??? You can't be serious dude. You are using a link to a picture of a benchmark for evidence. WOW! I vote that we replace this clown as the representative of AMD fanboys with Scientia. Scientia a real AMD fanboy who knows what the @$@#$ he is talking about.

This blog is so embarrassing for real AMD fans. Sharikou is a joke and can't be taken seriously anymore, just look at his resent post. For a real discussion about pervasive 64 bit computing please go to....

http://scientiasblog.blogspot.com/


HUH??? WTF is wrong with you sharikou? This story is in now way good news for AMD. You have no logic. I'm starting to see why more people are switching over to Scientia's blog. I think that you don't even care anymore what you write as long as it gets responses from intel fanboys. If you have the courage to post this then I urge the true Amd fans and the Intel fans to head over to Scientia's blog where the host is inteligent and is able to defend his points withour resorting to irrational garbage and ridiculous titles like "People rather wait for AMD than go Intel" What is that? Then you spew a couple of sentences about nothing. This is garbage that attract fanboys and provides no technical or financial information just a bunch of fanboys insulting each other. How about writing about technology and how pervasive 64bit is.

Still the same fanboy blog type. No change. Sctupidas is 100% AMD fanboy right thrugh. He just tries to hide it more. Still pathetic.

Scientia i keep on hearing you blabbing about about how these professional review sites are not doing the rite reviews. You have yet to show any application that mainstream consumers use that will hamper the FSB of kentfield. Could you

A) produce some evidence to back your claims instead of coming up with things out of ur a$$

B) do the kentfield tests yourself and show the professional site how proper benchmarking is done.

Mind you if you choose option B it has been tried before with sharikou when the woodcrests came in (still waiting on those results pretender) and he shut up after he could not back up any of his claims about the supposed benchmarking he was going to do.

---
So I've written quite a bit here that you've yet to address:) You're continuing to use the most obscure/weirdest 'chinks in Intel's armor' to support your pro AMD/anti Intel slant, saying Intel 'needs' Yonah because Merom is so hot yet fail to also address that Merom will ramp faster than Conroe. Oh yeah, 180 does not use 'AMD fanbois say this, look they're wrong' and even Sharikou rarely does that. The way I see it..
You: Hey, he said something not nice about AMD! How could he have said that? I think I'll tell the whole world once I saw this person saying something mean about AMD, even if it makes no sense to use one person's isolate view to back up my belief that all[or 'many'] Forumzers are evil:/

ashenman said...

"So silly Scientia, to address me while I'm anonymous on 180 yet ignore me on your own blog:/"

Thanksgiving break? Maybe he's just trying to actually take a break from blogger...

Anonymous said...

K, if you say so, but he's responded before/after Thanksgiving elsewhere:D BTW, I know that lately I've become much more sarcastic/etc and I'll try to tone it down:)

Anonymous said...

Hehe, remember all that bickering over something that was bound to release in a couple weeks anyhow over the FSB starving? Well I was wrong in my assessment that Quad FX would have some lead in some multithreaded apps. :D I wonder why AMD didn't bother with 65 nm 2.6 FXs.. And seeing how this underhyped..thing came out to be, I wonder how 65 nm will pan out.