Friday, May 26, 2006

Intel finally making some moves as I proposed

Back in 2005, I made the following suggestions to Intel, which I sent to Intel execs via email:

It seems that the only way out is for INTEL to drop the most favoured treatment for DELL and increase prices cross the board, and make prices the same for everyone, making it possible for every one (instead of just DELL) to profit in the INTEL space and create a more balanced INTEL market.

My suggestions to INTEL:

0) Face the reality that AMD will take 25% share.

1) Abolish the volume based or AMD quota based rebate system.

2) Establish an uniform pricing scheme, set the same equal global price for everyone, just like McDonalds sell cheeseburgers at the same price. Get rid of all discounts. Such as an uniform pricing system will make small whitebox players competitive against big guys like DELL and foster a broader market presence.

3) Reduce the production of low end chips, this will lead to higher ASP.


I was trying to find a way for Intel to survive the AMD onslaught. Later, I further explained why Intel should cut rebates to Dell. I projected that Dell had to go AMD by the end of 2Q06.

Now, it's reported that Intel is finally doing 1), 2) above. However, Intel failed to follow my advice to reduce production in anticipation of AMD's capacity increase and market share gain. As a result, Intel's inventory piled up to the neck. This has led to massive oversupply of Intel's legacy chips and Intel's pricing collapse. Intel is now trying to cut prices by up to 60% to gain back 5% of unit share, which is kind of wishful thinking. Intel probably thought that they could halt sales of AMD64 by hyping Conroe. Of course, that was naive. People still need computers, so they go AMD64, because Intel told them Pentium 4 is indeed junk. Intel's slashing of the P4 price further validates the notion that P4 is junk. The continuing steep price reduction will only cause buyers unwilling to pay whatever reduced prices today, in expectation of further price drop as Conroe launch quickly approaching. As customers wait, Intel sees demand drop; to stimulate demand, Intel cuts prices even more, causing more customers to wait for further price cut. The cycle continues. The pricing collapse also causes Intel to lose credibility and customer confidence. Those who bought Intel earlier definitely regret. This may even push more customers to AMD, as they feel AMD's prices are more stable and fairer. As I observed here, people are not buying Intel's Pentium Ds, even though their prices have already been cut to rock bottom levels.

I projected operating losses for Intel from 3Q06 onward (2Q06 GAAP loss is also pretty much in the bag due to impairment of goodwill). As the launch of Conroe will cause total collapse of pricing on 80% of Intel's chips. Intel may be able to ramp Conroe to 40% of its volume in 1Q07. However, by then, AMD's superior 65nm CPUs will start to flood the market, Conroe will again become obsolete.

Intel has made the following irreversible errors that ruined its only chance of survival:

1) Intel hyped Conroe too much and too early
2) Intel failed to reduce production to hike price and keep revenue and profitability constant
3) Intel triggered a cycle of domino effect type pricing collapse of its CPUs
4) Intel has squandered its cash, leaving it out of ammo in the death struggle against AMD

Inventory write-downs, pricing-collapses, goodwill impairments, cash burn will cause a vicious cycle of spiraling down. Fighting capacity and morale will get lower and lower...

The story has a sad end for Intel.

34 Comments:

Blogger netrama said...

I am sure the morale among the Intel employees is already low. I can't help , but feel sad about their situation.
It is an evil truth that every one in the ship has to suffer due to the blind whims (GHz race,"we will kill AMD") and fancies (market share) of a few.

11:32 AM, May 26, 2006  
Blogger Sharikou, Ph. D said...

Conroe could have saved Intel. Had Intel
1) maintained radio silence on Conroe
2) reduced production to cause 5% shortage of CPU supply and 15% price hike
3) established uniform pricing to foster the whitebox market
4) generated revenue and profit growth despite limited unit share loss
5) moved capacity to accumlate Conroe parts
6) conserved cash
7) initiated massive surprise attack in 3Q06

AMD would have to scramble to react. Intel would be able to keep its forces in high morale and continue to battle on the next architecture.

Now the situation is Intel is goning to suffer massive losses in the quarters ahead despite Conroe's perceived strength. By the time Conroe reaches critical mass, AMD will be out with super frags, causing Intel unable to profit and regroup.

12:11 PM, May 26, 2006  
Anonymous Anonymous said...

"NGMA" - The Phantom Menace ?

3:55 PM, May 26, 2006  
Anonymous Edward said...

the problem is Intel changed CEO, and the new guy wants to do something big.

that had always led to problems in history.

6:18 PM, May 26, 2006  
Anonymous Anonymous said...

Couldn't Intel conserve some cash by killing off the Itanic? Ha ha.

9:59 PM, May 26, 2006  
Anonymous Anonymous said...

So mush of the fact that 64 bit doesn't matter.

http://www.theinquirer.net/?article=31989

It is the action of a desperate man. But cutting prices so much in so short time never yiels to something good. And what about an image that INTEL displays?? It's like saying: our products are not good, and we were overcharging them, but now you can have them for free?!
very, very sloppy.

1:30 AM, May 27, 2006  
Anonymous Anonymous said...

sharikou,
Maybe what you logical and sensible. However, What Intel done so far is predictable.

Think of it. If Paul did exactly what you said, both revenue and earning will come down slowly and Intel survives. The CEO of the Intel, Paul, does not look good. He would rather gamble on anythings by any means.(selling tomorrow products today, NGMA, cutting price etc..). If he can turn it back, Paul will be the hero. If not, who cares (stockholders should).

In a nutcell, Paul is doing the right thing to him(may not to the stockholders). then again, Paul is not the frist CEO in history acts like this.

The question is : If you are the Intel stockholders, do you want to dance with Paul?

3:52 AM, May 27, 2006  
Anonymous muziqaz said...

sharikou, what's your take on this:

http://www.theinquirer.net/?article=31994

because english is not my native language, so it is very interesting for me to trie to understand what is spring struck Charlie blathering about.
Besides Sharikou, if You say, that now one is buying underpriced pentiums, so why people would have been buying those chips if intel would have stayed with the same prices?
Though in other hand, Conroe messed everything up.

12:08 PM, May 27, 2006  
Blogger Sharikou, Ph. D said...

sharikou, what's your take on this:


I have concluded that a while ago that Conroe only has an advantage in 32 bit cache sensitive single threaded applications. Just wait two months and see.

7:31 PM, May 27, 2006  
Anonymous Anonymous said...

Erm, and what app is not cache sensitive? Notepad? SuperPi is extreme example of Conroe benefitting from cache, but Hexus and XS have shown that it has completely no problems dealing with faster clocked K8.

Conroe has no problems against K8 at any workload. Sciencemark is uber optimized for K8 code and yet Conroe can keep up. Am i surprised? Hell no with massive theoretical IPC advantage it's all about making use of it :)

Average Joe will buy entry AM2 system let's say X2 4000 and DDR2 533 and Conroe 2.4 with same DDR2 memory. I am certain that Intel will outperform AM2 at least 50%, cause AMD is taking HUGE hit due to DDR533.
Tho

1:30 AM, May 28, 2006  
Anonymous Anonymous said...

no one (in the consumer market) cares about the maker of the processor. else amd would have been bancrupt a long time ago (when their processor weren't competitive with that of intel).
it's all about the price...

3:34 AM, May 28, 2006  
Blogger Sharikou, Ph. D said...

what app is not cache sensitive?

Data intensive applications are not senstitive to cache. DivX 6.1 and WME9 show almost zero performance increase when cache increases. Games are more sensitive to cache sizes. Increasing Athlon 64 939's cache from 512K to 1MB leads to up to 8% increase in gaming performance.

9:04 AM, May 28, 2006  
Anonymous Anonymous said...

Data intensive applications are not senstitive to cache.

It depends on memory access pattern. DivX 6.1 and WME9 have "streaming" patterns, usually operating on fixed size blocks of data. Those are easy to prefetch and usually fit in L2 cache unless it's extremely small (read celeron, sempron). But most other "data" intensive apps - games, data archiving, 3d rendering, data mining, application servers have way less uniform access patterns and benefit from enlarged L2 cache a lot.

Imho this Intel generation is all about brutal strenght:
* 65NM process enables 4M L2 cache
* brutal execution core
* even more brutal SSE execution capabilities.

And ironically - DivX 6.1 and WME9 are workloads where K8 is beaten badly by Conroe. If you look at Hexus Conroe test and Divx results:

AM2 FX62 : 164s
E6700 : 125s

AMD seriuosly needs to rush K8L and to cut the prices in half.

10:10 AM, May 28, 2006  
Anonymous Anonymous said...

the tumbling of intel will hurt all processor manufacturers. if intel lowers its prices, all others have to follow killing the margins...

12:21 PM, May 28, 2006  
Blogger Mad Mod Mike said...

Link

It appears that Intel hasn't learned and wishes to claim they are still in the race with AMD. It's sad.

1:22 PM, May 28, 2006  
Anonymous Anonymous said...

Is anything besides that large cache actually making conroe faster than K8? such as Micro-opps fusion or whatever it is? Because it looks like once each core only has 2 megs of cache to work with when multiple task are run, conroe is not faster than K8.

So I guess this is just outmanufactoring by intel, not supperior design. Then again, nobody cares, because that super pi score sure looks mightily impressive

10:42 PM, May 28, 2006  
Anonymous Anonymous said...

the importance of cache size is very much depend on speed (latency) of memory.The lower latency we have, the lower cache size we need, and vice versus. AMD, unlike Intel, much better utilize advances in RAM development. Now when Corsair started production low latency DDR2 800 (CL3) and DDR2 1000 (CL4), AMD gained much more than Intel. In 4-6 months as we will see CL2, big cache of Conroe will turn in a burden.

1:34 AM, May 29, 2006  
Anonymous Anonymous said...

Yes Sharikou, I agree with you ! Intel goes down day by day. Intel is doomed. Go go go AMD. (Rathor)

3:11 AM, May 29, 2006  
Anonymous Anonymous said...

"I am certain that Intel will outperform AM2 at least 50%, cause AMD is taking HUGE hit due to DDR533."
hi intel fanboy.. how's your "con-roe" day dreaming going?
50%? please show me a benchmark that shows Conroe being 50$ faster than AMD.
according to this site,
"http://www.legitreviews.com/article/347/3/"
the problem with IPC is that you need a decent RAM to really see significant benefits.

8:07 AM, May 29, 2006  
Anonymous Anonymous said...

Sharikou, would it really kill you to admit Intel has come out with a great chip? Give credit where credit is due.

9:22 AM, May 29, 2006  
Blogger Sharikou, Ph. D said...

Hexus Conroe test and Divx results

Hexus? A bunch of kids. Do they know how to do arithmatic?
Read this benchmark for Woodcrest and Opteron. Remember, Woodcrest is just Conroe with a different label.

1:41 PM, May 29, 2006  
Blogger Sharikou, Ph. D said...

would it really kill you to admit Intel has come out with a great chip? Give credit where credit is due.

All Intel has to do is act like with some ethics, stop the cheating and learn how to design a chip.

1:43 PM, May 29, 2006  
Anonymous Anonymous said...

to edward. intel's change of ceo was already known before it was going to happen. i believe intel has strict retirement policies in which once you hit a certain age, you must retire and the ceo was getting to that point, it was not like the ceo just stepped down with no warning....

2:54 PM, May 29, 2006  
Anonymous Anonymous said...

since to claim to have a ph.d., and make it seem like you know how to make a better x86 processor than intel, THEN SHOW US. seems like the ones that really understand something-do it, the ones that can't do it-teach it, the ones that can't teach it-manage it or just talk about it (this would be you in the LAST category). since you are SO brilliant, why don't you get a job at intel to show them how to make a cpu "the right way" or work at amd? or even start a new x86 processor company? it is obvious that ANY x86 processor designed by you would be better in: performance, power consumption, marketing, etc. why would anyone want a processor from amd or intel when they could get a better one from you?

3:05 PM, May 29, 2006  
Anonymous Edward said...

"intel's change of ceo was already known before it was going to happen. i believe intel has strict retirement policies in which once you hit a certain age, you must retire and the ceo was getting to that point"

What are you talking about? You completely missed what I was saying. I DID NOT say Intel changed CEO due to poor performance or any reason. I DID NOT say Intel changed CEO without warning. What I said was that the new guy wanted to do something big could potentially lead to problems.

Paul Otellini changed Intel's logo, Intel's slogan, Intel's marketing strategy & ethics (to even lower grade), and maybe also Intel's corporate structure, he REALLY tries to do a lot. It seems he wanted to show the poor investors that HIS changes made Intel profitable again (hopefully). But IMO, he is sacrificing Intel's morality & integrity for that, and that's BS. I'd rather Intel spend more resources at engineering to make a better Core 2 Duo (rather than benchmarketing it to be!)

8:24 PM, May 29, 2006  
Anonymous Edward said...

"And ironically - DivX 6.1 and WME9 are workloads where K8 is beaten badly by Conroe."

Seriously, a few lines of change in how instructions are ordered could lead to that much performance difference, especially for mmedia programs.

That's why we sometimes see A64 bashing P-M in mp3 but P-M bashing A64 in WME9 (or vice versa). If you think of it, these two programs are doing about the same stuff - lossy audio compression.

8:34 PM, May 29, 2006  
Anonymous Edward said...

"Is anything besides that large cache actually making conroe faster than K8? such as Micro-opps fusion or whatever it is?"

It's funny to see the Intel fanboys touting Conroe's micro-ops fusion simply because they don't know much enough about Pentium-M, which already contains the feature (btw, so does AMD's Athlon64).

Conroe's most significant advantages:
(1)4-issue pipeline (versus 3-issue in P-M and A64), and for that matter increased instruction window & retirement units.
(2)Large 4M shared L2 cache (incl. direct communication between L1 caches)
(3)Better prefetch (result of better memory address disambiguation)

In average each above could contribute 5% performance increase in real programs (compared to P-M). Some of the additional boosts we see from those benchmarketing sites, if they are precise (ie. not lying), could come from the following:
* benchmark fits into L2 cache
* software optimization
* aggressive memory timing

8:59 PM, May 29, 2006  
Anonymous Anonymous said...

Theres no way on earth that sharikou would make it past a phone screen if intel was hiring even if he did intel grills people for a whole day before it hires them. I think sharikou would opt out anyways.

Now secondly im sure my comments are gonna be censored since he only posts blogs of intel bashers anything pro intel is censored.

9:51 PM, May 29, 2006  
Anonymous Anonymous said...

"Now secondly im sure my comments are gonna be censored since he only posts blogs of intel bashers anything pro intel is censored."
that's because you intel fanboys have been posting meaningless (to the extreme) comments on this blog. sharikou is basically stating his opinion, and back them up with good sources.
and what do you guys do?

10:45 PM, May 29, 2006  
Anonymous Anonymous said...

One thing I have noticed in reading the posts here is that when you talk about "Conroe" beaing AMD you always talk about how much cache it has, and how it only wins because of its cache. Well I am not a computer engineer, but if you have a product that is able to compete because of a larger cache then how can you people bash the design.

Some one else pointed out that a larger cache is there because of latency issues. Again, how can you bash the design, all Intel is doing is working with what they have, right?

7:17 AM, May 30, 2006  
Anonymous Anonymous said...

"Some one else pointed out that a larger cache is there because of latency issues. Again, how can you bash the design, all Intel is doing is working with what they have, right?"
the problem with intel and conroe's large cache is intel kept on benchmarketing, fooling everyone about how conroe has an advantage of over 30% over AMD's fastest offering. in truth, conroe's large cache may only significantly benefit games (that's why intel used them in IDF), some synthetic apps, and some cache sensitive apps. if put under server apps, NGMA can't even stand a chance (that's why intel changed x86 on TPC site to 64bit on intel's site, to stay competitive).

9:14 AM, May 30, 2006  
Anonymous Anonymous said...

Why does this website remind me of listening to the Iraqi Information minister?

...The Americans are nowhere near Baghdad...Intel has spent all its cash and is defenseless against the AMD onslaught. Whatever.

8:00 PM, May 30, 2006  
Anonymous Anonymous said...

"Why does this website remind me of listening to the Iraqi Information minister? "
iraqi info minister had nothing to support his claims. sharikou, however, backed up his claims by numerous sources. he may not always be right, but he had done his hw.

9:00 AM, May 31, 2006  
Anonymous Anonymous said...

But Sharikou also blatently dismisses sources that condradict his conclusions. For example, Sharikou long ago labelled Anandtech, one of the largest and most reputable tech sites on the internet, as an "intel pumper". Even in this very comments section, Sharikou just throws out the Hexus benchmarks a second thought.

6:53 AM, June 01, 2006  

Post a Comment

Links to this post:

Create a Link

<< Home