I admire Hector Ruiz
The guy is a great military strategist. The massive attack with $149 X2 3800+ and even lower X2 3600+ coupled with DELL's massive volume capability will truly deny Intel the oppurtunity to clear its legacy inventory and result in massive operating losses for Intel. AMD is gonna ramp FAB36 and Chartered FAB7 like crazy. The goal is clear, deny Intel from selling Pentium Ds at a price above $90 at all costs. Since an X2 3800+ frags Pentium D XE 965 at much lower power consumption, AMD's move will make the Pentium Ds (which will remain 75% of Intel's volume) 100% unwanted.
I declare Intel clinically dead in 5 qaurters instead of the previous 7 I projected.
There are some hand waving arguments by some people on AMD's capacity, which are clueless. AMD's chip die sizes are always similar to Intel's. The Pentium D and Conroe's die sizes are about 160mm^2 at 65nm due to their 2x2MB cache. AMD's X2s have die size of about 180mm^3. Assume AMD is now at 25% X2 and 75% single core, the average die size is 117mm^2. Suppose AMD starts to produce 65% X2s and 35% single cores, the average die size is also 158mm^2. The increase in average die area is 35%. However, look at page 5 of this slide, you see AMD's wafer output increases by 50% in mid 3Q06, add another 10% increase from Chartered, AMD's chip unit output will increase by a factor of 1.6/1.35= 118.5% by mid 3Q06. By 2Q07, AMD's total wafer output doubles, however, because FAB36 is 50% converted to 65nm, the total equivalent output will increase to 2.5x of current level, leading to 2.5/1.35 = 1.85x current unit volume, throwing in some more from Chartered, 2x output is an easy goal. That's not all, AMD's capacity will go straight up from 2Q07 till 2009.
It's impossible for Intel to grab more market share, since an X2 3800+ frags a Pentium XE 965, so Intel will continue to lose market share quickly. Intel's ASP was $150, and AMD seems to have deliberately priced X2 3800+ at $149, below Intel's ASP. For Intel to maintain its ASP, for every chip it sells at $100, it has to sell another at $200, which is impossible, because Conroe's volume is limited. Lower units coupled with lower ASP, will lead to massive operating losses for Intel.
For AMD, its ASP now is below $100, so every chip it sells at above $100 is a winner. Actually, if AMD can simply sell one X2 3800+ at $149 plus one Sempron at $51, its ASP will hit $100 and an increase from current levels.
posted by Sharikou, Ph. D. @ 7/19/2006 11:15:00 AM
41 Comments:
Ok I will admit I am only looking at one review, but how do you comeup with...
"Since an X2 3800+ frags Pentium D XE 965 at much lower power consumption"
I will not deny price/power/performance, but frags Intel 965, please link to the article you read that in please.
The one I was looking at had Intel 965 winning most of the benchmarks, see below.
Q4 = Intel 965 7% advantage
BF2 = AMD X2 3800+ 1% advantage
HL2 = Intel 965 1% advantage
F.E.A.R. = Intel 965 4% advantage
ROL = Intel 965 17% advantage
Oblivion = AMD X2 3800+ 3% advantage
I am not trying to justify paying the price of a 965, but just point out that the Dr, who is usually not so biased, is not looking at the numbers, he is just trying to make AMD look good.
Also for a very good breakdown of this pricewar, I urge you to take a non biased look at this article.
I based on this
benchmark between X2 3800+ and XE 965
Office; X2 3800+ 1% faster
Mozilla: X2 3800+ 22% faster
Mozilla +WME: X2 3800+ 15.9% faster
In other tests, both are quite close.
My bad I was only looking at gaming.
Thanks for responding.
Doctor Doctor anyone home.
It takes quarters to ramp from 0 to full production.
Lets take Charter; there is NO way AMD nor Charter would have ramped from 0 to thousands of wafers a week even if th fab was already at that capacity without a pilot run to validate the process health, yield, perforance etc. etc. Once validated you would turn it on. You just don't turn on a factory overnight.
Second AMD's 300mm factory aint' coming up overnight. It takes quarters to ramp a factory from 0 to thousands of wafers / week and then another quarter for you to see it in the OEM's hands. So AMD won't be giving anyone icnremental huge volumes at low price any sooner then the Conroe/Merom ramp lands in everyone's lap.
Its tough.. there is on quick, simple, or fast way around AMD's capacity issue.
Actually Hector was an idiot. He and Jerry should have seen this two years ago and should have accelerated their 300mm ramp. Imagine if AMd had done that they would have had their 300mm factory cranking 9 months ago when they were still competing against Netbust. Now.. hard as you try COnroe and Merom are in everyones mind and everyone knows INTEL has 3 300mm 65nm factories cranking.
Dell.. You want a Conroe/Merom, a dirt cheap Pentium D.. INTEL has it all. Want a dirt cheap AMD chip.. well I can sell a few to you but you gotta wait till Q1'07...
I don't think so.
The Doctor
Dude, you need to learn to make your ridiculous projections last longer. Was it GhostBusters where the psychic predicted the end of the world was in like 20 days?
Your silly little predictions about Intel's demise will look pretty foolish in Q1/Q2 of 2007. You should have said sometime in 2008 so you would at least have a year+ of the little credibility you have with the other kooks.
I would like you to comment on this article.
It would seem that the lowend Conroe is a great over clocker, and at a price of just $183, it sure would be hard to pass up.
It may not be comparable to 4x4, but the price is great.
Thnaks
If there are many people in the world as thick-headed as the purchasers at my company, Intel will still be able to clean out it's inventory of P4's and P4-based Xeons, even if it's at a very low profit.
That's why I proposed the only possible way for Intel to survie and avoid BK: reduce production and raise prices.
But with a pricing crash, no matter how many Intel fans are there, they won't be able to save Intel, because Intel can't say "Ok, you are a fanboy, pay $200 more". Once the price is slahsed, everyone pays the same low price, fanboy or not. Furthermore, slashing the price won't increase the number of Intel fanboys significantly.
The net result is thus inevitable: lower units and substantially lower price--massive operating losses for Intel.
It takes quarters to ramp from 0 to full production.
The AMD chart is for wafer outs! Not for wafer starts. Finished Chartered CPUs are selling right now. Finished FAB36 CPUs started selling in March 06.
It would seem that the lowend Conroe is a great over clocker, and at a price of just $183, it sure would be hard to pass up.
It doesn't matter at all. 75% of Intel's units will be Pentium Ds. Since X2 3800+ will be sold at $149, and there will be a X2 3600+ at $129. Intel can't sell P4/P4Ds at higher than $100 ASP. So Intel's ASP will be
$100 * 0.75 + $x * 0.25
where $x will be Conroe ASP.
If you want to keep a $150 ASP, then $x must be $75 /0.25 = $300
So unless Intel sell Conroe at an ASP of $300, it will see an overall ASP drop.
But you can clock it by 400 mhz at stock voltages so you can get a x2 4600 with wich demolishes the pentium ee by a mile
Of course they will have to lower the prices when they know they can't compete on performance/price as the scenario was before the price cuts.
That's the exact same thing Intel did with their Pentium 4 series when A64 754/939 came out.
Let's say the AM2 5000+ performs 15% worse than Conroe E6700 overall. Then AMD has to lower the price of the AM2 5000+ by more than 15% under the price of E6700 to stay compeditive.
It's just a simple example of the facts of healthy competition.
What we have in this blog is an arrogance that far surpasses that of Intel.
AMD is shipping nothing *today* that compares with *today's* Intel chips (Conroe, Woodcrest).
It is sheer hubris to keep up this grandstanding.
AMD's micro-architecture team has failed to produce any improvements in 5 years. In fact, they have taken steps backwards with the AM2 performance dud.
AMD's 4x4 effort just made AMD's AM2 "future proof" commitment into hot fresh slag.
AMD has raised Opteron prices instead of lowering them to compete with Woodcrest.
AMD has announced vaporware solutions (4x4) instead of better chips.
In the objective world, AMD is not making more money than Intel, doesn't have as much cash, and has nothing of the infrastructure. So there are no innovations from AMD like FB-DIMM. Instead you have the performance bog of AM2/DDR2 without the on-chip cache to make up for the memory latencies.
Basically AMD is a bunch of stupid guys who got lucky with one good design. Whomever did that design obviously left AMD a long time ago and like a bunch of dumb monkeys, AMD has not known what to do since that time. It is very close to game over for these monkeys.
Is this not a repetition of the end of the second world war,the fuhrer waving on his soldiers for the last thrust,waiting for the magical weapon, while the allies are banging on their doorsteps.
Some interesting comnments but some of them just don't make any sense:
"Intel's ASP was $150, and AMD seems to have deliberately priced X2 3800+ at $149"
Why would you price a specific chip against an ASP, if AMD is doing that they are truly stupid. They should be pricing the chip against competitive products.
"For AMD, its ASP now is below $100, so every chip it sells at above $100 is a winner."
Actually that's not true either - die size, yield, bin split for the particular chip are all important factors. If AMD starts selling a bunch more FX62 chips at $110 (above your ASP #) does that help or hurt revenues?
"By 2Q07, AMD's total wafer output doubles..."
I think you mean die output? Converting between 90nm and 65nm will not change WAFER output but it will impact die output; it actually may hurt WAFER output a bit as there are more process steps (like an extra level of metallization) over the previous technology node. There is also some short term capacity reduction during the transition from one technology node to another (each 90nm tool needs to be qualified on 65nm process and then some tools may need HW upgrades).
"Since X2 3800+ will be sold at $149, and there will be a X2 3600+ at $129. Intel can't sell P4/P4Ds at higher than $100 ASP."
By this logic FX62 needs to be priced at or below an E6600 (~$320).
While one would think pricing is only set by performance levels, in the real world this is not the only factor for pricing.
Some interesting comnments but some of them just don't make any sense
I don't have time to explain. But my analysis was obvious to someone with some level of IQ. Maybe others can help you.
"I don't have time to explain. But my analysis was obvious to someone with some level of IQ. Maybe others can help you."
I don't have time to explain or can't come up with another BS explanation. Good to see your objectivity shining through yet again.
Have you ever mentioned what your PhD is in and what university it was from?
"It doesn't matter at all. 75% of Intel's units will be Pentium Ds. Since X2 3800+ will be sold at $149, and there will be a X2 3600+ at $129. Intel can't sell P4/P4Ds at higher than $100 ASP. So Intel's ASP will be $100 * 0.75 + $x * 0.25"
Are you factoring in mobile and server too or just desktop? Is the ASP you are quoting desktop only, not sure where you got the 150 from?
You notice how lately it has been nothing but Intel Fanmonkeys commenting? I think they all horde together because they don't want people knowing how bad Intel is and they NEED to somehow disprove the truth (which in itself is retarded) so that's why you get 20 posts from a Intel fanmonkey blabbing the same thing (that makes no sense btw) and than 1 AMD Intelligent Person proves in a few paragraphs "Sorry, you're wrong".
Funny the way things go, eh mates?
Wrote the anonymous coward ,
"Basically AMD is a bunch of stupid guys who got lucky with one good design. Whomever did that design obviously left AMD a long time ago and like a bunch of dumb monkeys, AMD has not known what to do since that time. It is very close to game over for these monkeys."
What we are witnessing here is the collective wisdom of Intel fanboys , being Demonstrated in uglier fashion.
Although I'm proud of the job well done by Hector Ruiz thus far, it is my belief that:
1) AMD did not expect Intel to bounce back so soon, so strong (6-12 months early).
2) AMD is mangled into this price war they didn't want or need.
Will Hector and Co. be able to find a solution to this little problem? How long will this problem last?
My guess is yes and for the next 6 to 9 months. After that, I truly believe AMD will pull out some technology that we will all be amazed by. No I'm not pulling an Intel fanboyism statement here. I'm just relying on what Hector has said before in an interview (smmarized in my own words):
"Intel believes we were just sitting there doing nothing while 'they' were improving their technology (Core). [...] We will be introducing some jaw-dropping inovations in the near future."
This guy is credible and trustworthy. I'm expecting no less than a new inovation soon!
Mad mod mike; ho stick your head in some more glue and feathers covered with AMD and nVidia tags
1) AMD did not expect Intel to bounce back so soon, so strong (6-12 months early).
2) AMD is mangled into this price war they didn't want or need.
In a war, you find a way to turn your disadvantage into advantage. Hector did exactly that. From a long term point of view, Intel has already made fatal mistakes, and will ultimately BK because of those mistakes. AMD on the other hand has responded with a clear head. If you go back and read my blog entries, you can see AMD's moves have mostly been predicted by me--I was simply trying to find the optimal solutions for both Intel and AMD from a military strategy point of view. However, Intel retards failed to act coherently.
Judging from Intel's 2Q06 report, I can safely say Intel is finished. By mid-2008, you will see Intel at about 25% of its current size, while AMD takes the dominant position. By 2009, Intel will be reduced to a very small player.
Sharikou: "I based on this
benchmark between X2 3800+ and XE 965
Office; X2 3800+ 1% faster
Mozilla: X2 3800+ 22% faster
Mozilla +WME: X2 3800+ 15.9% faster
In other tests, both are quite close."
Just so I understand for future reference when a chip outperforms another on Mozilla and all others "are quite" close (which isn't true - see below) that means the chip "frags" the other?
Or is is the 1% office advantage that leads to the "frag"ing? Apparently when the 965 leads in something it is close, if the 3800 has a +1% advantage that is a whole different story?
Please click on the "next" button a few times on the link YOU SUPPLIED, you'll see the 3800 is often behind the 965. It's good to see you aren't cherry-picking the benchmarks to support your point of view.
I hope you post this opposing point of view.
"Mad mod mike; ho stick your head in some more glue and feathers covered with AMD and nVidia tags"
Someone in this blog mentioned earlier " FOR SOME INTEL IS LIKE A RELIGION " .
In my opinion it should be"
For some Intel is like a religion,
and its a BAD RELIGION.
Dr. Ruiz is indeed a strong leader, but actually what 'changed' AMD were more Intel's Pentium and Itanium.
Pentium forced AMD to make K5/K6, which are the first two independent microarchitectures from AMD. Itanium destroyed DEC Alpha, and allowed AMD to acquire a world-class design team from it for Athlon.
Also noted is AMD's success in merging the different talents and resources from the three groups (K5, Nx586, and part of the previous Alpha team). Not many acquirement in this industry worked as well.
Of course, it took lots of efforts and pain for AMD to climb up and retain its 20% market share (up until recently). Honestly every step of it since 1995 is a step upward. That's not counting the poor years between 2001 to 2003, during which P4's Megahertz hype successfully hindered the sales of Athlon & Athlon XP.
A Intel fanboy who doesn't know much in-depth history has his reason to look down upon this company. After all, until recently, AMD made almost no money, and took only 10-20% marketshare. It is partially true that "AMD lived because Intel wanted it to live," at least until the K6 (1997).
However, reading the history more properly one will find that AMD really started its own x86 design around 1994; it came up with an admirable one in only 5 years (Athlon). This is accomplished even under Intel's "evil" monopolistic marketing tactics. Considering Intel's crappy CPU designs in the first twelve years (from 8088 up to 80286), what AMD has done is really remarkable.
Well, lots of Intel fanboys, many of whom also intel shareholders, won't admit that. They will point to Intel's past glory. IMO, they are just blind to the current and the future.
Sharikou, you mention 75% of all intel units will be pentium D. Did you include mobile chips in your numbers? As I understand, laptop sales are growing fast, and are close to overtaking desktop parts. And most of these are pentium m/core duo chips. So intel still has a very strong position here (although tough competition is coming within half a year).
it will be really good if you go join AMD. with you prediction, planning skill, coupled with the PhD, AMD will sure benifit from them and become the #1 company in the world.
I believe all of us that visited your blog will surely support you on that. You see, those that believe you, sure thinks that you would greatly help AMD, and those that do not like AMD and do not believe you, surely think that you would somehow destroy AMD. Go, my friend, you can work in the company that you like, meet with Hector occasionally and prove those that do not believe in your predition/planning/techno savvy people to be WRONG.
I'll be the first to support you here! Go, Sharikou, go go go.
Keep in mind that Intel has known about this problem for a long time. It was first revealed during the 1Q06 Bryant and Otellini reports that Q2 would be a very rough quarter for Intel. They expected this. They prepared for this. They are not "fragged" bcause of this.
Think about it: when did Intel start working on its "efficiency drive"? I'll tell you: May 1. Why did they start working on this efficiency drive? Because they knew they were going to get hosed in Q2.
The problem I have with your claims of an Intel on the verge of Bankruptcy is that it simply lacks truth, factual support, and common sense. The one time I thought Intel was in serious trouble was when the middle of 2005 when their roadmap was the weakest, and AMD's lineup outperformed theirs at every turn. Guess what though, AMD still couldn't compete, even with better products! (blame it on a poor sales department, a bad reputation, whatever you want) That's when the lawsuit was filed, and that's when I knew AMD was in trouble. Why? Because with that legal action they basically said: "Even with better products we still cannot compete." And now AMD has an admittedly worse roadmap than Intel. Do you honestly expect a company that couldn't compete with better products to be able to compete with inferior products?
Now, I wouldn't go so far as to give a timeline for either companies demise. Both will likely be around for a long time, which has been a blessing for the market. I would highly recommend however, that you stop this incessant banter about Intel folding in a year. It just makes you look like an idiot.
Judging from Intel's 2Q06 report, I can safely say Intel is finished. By mid-2008, you will see Intel at about 25% of its current size, while AMD takes the dominant position. By 2009, Intel will be reduced to a very small player.
No, intel will BK in 5 quarters! didn't you say so?? *it is actually 5 quarters - 2 days), i'm keeping score for you. But can you restate it again, because some of your post said, 7, then 5 and some said 2 .. i'm losing track now
No, intel will BK in 5 quarters! didn't you say so??
I was expecting a chapter 11 BK, not chapter 7.
well its impossible for AMD to gain back market share too, because E6300 completely frags X2 3800+, not to mention E6700 trouts FX-62 (and very likely 64 / 66) around.
http://www.maxitmag.com/testing-bay/processors/intel-1.86ghz-core-2-duo-e6300-review/
"That's when the lawsuit was filed, and that's when I knew AMD was in trouble. Why? Because with that legal action they basically said: "Even with better products we still cannot compete."
That's the reason Hector decided to file the Anti-trust lawsuit!! In spite of Intel having a weaker product, the used their muscle, coersion, and monopoly to sell people like you inferior products at a very high cost.
This my friend, is why monopolies are illegal.
If it weren't for AMD, you'd be paying up the wazoo for P4's still... ever think about that?
The reason why intel ceased further development of the netburst architecture, and came up with conroe, is to be able to withstand future competition with AMD. Using 45 nm process, netburst would probably scale much higher; but it would not be competitive with what AMD will come up with soon. So even if FX62 is beaten on all fronts right now, just wait a few months and see what happens then. You will be surprised.
"just wait a few months and see what happens then. You will be surprised."
Could you let us in on your inside information?
Just take a good look at the Phil Hester presentation on the AMD website.
http://www.amd.com/us-en/Corporate/InvestorRelations/0,,51_306_14047,00.html
And more here:
http://www.laptoplogic.com/news/detail.php?id=960
Ok, I am not trying to critisize your links, but there is nothing definitive in what you linked too.
"The end result, regardless of our speculation, is a CPU that will be extremely competitive with Intel's upcoming Merom processor."
To me the problem is neither Merom or Bulldozer are out yet. We don't know how they will "real world" compare, also how much market share will AMD lose, Merom has been pulled forward (suposedly) and Bulldozer is still at least half a year away?
The AMD presentation, well thats all it is, until there is actually a processor being benched its just that, a presentation. I will not try to say that Intel is great or anything like that, but all in all they benchmarked a chip(was it 3 or 4 months) prior to release, and it has achieved what they said +/-.
Until AMD has physical samples in someones hands its not real. I have no doubt they will be back, better for you and me when they do, but they need to respond to C2D, and they need to do it soon, 4x4 is mainly for heavily multithreaded applications, and thats just going to be ahead of its time.
Thanks for the links though.
You are correct, there is nothing definitive in it.
But Dell just "broke up" its exclusive deal with Intel, and I think they must have a very good reason to do so, and to do so now. AMD is also desperately trying to expand fab capacity, even by ordering more expensive Chartered chips.
On the performance and performance/Watt side however, I have trust when looking at the technical specs.
The next two quarters will be uncertain, but I am confident C2D will not be that much of a problem. Most people only upgrade once in a couple of years (I got my x2 3800+ in january to replace an old Pentium III 800eb), and both the athlon x2 / C2D and even pentium D will offer a terrific improvement for those people. Only gamers need to upgrade once every two year. I am sure a lot of people will upgrade to a dual core cpu within the next quarters, due to the lower prices. AMD is prepared to offer a lot of relatively cheap 512 kB x2 cpu.
"Anonymous said...
What we have in this blog is an arrogance that far surpasses that of Intel.
AMD is shipping nothing *today* that compares with *today's* Intel chips (Conroe, Woodcrest).
It is sheer hubris to keep up this grandstanding."
mind taking a photo of you holding your recently bought conroe? :>
So, did you get to screw his ass or did he do yours?
Post a Comment
<< Home