Why isn’t AMD Promoting its 90nm CPUs More?!?

30 Nov, 2004 at 9:46pm ET Article published in Tech

Overclockers.com poses and offers it’s opinion on this question

“With all due respect, we think AMD has very good reason for staying quiet now and is quite wise in doing so, for now.

This article will explain why, again, with some new supporting information, and in the end, some new developments that ought to eventually be very good news for AMD fans and give a good indicator as to when the best time to go to Hammer ought to be for most. ”

AMD has managed to convince an obviously not-too-inquisitive media that it has succeeded where Intel hasn’t and can make cool-running 90nm chips.
What has escaped the media’s attention is that AMD is making relatively slow, cool-running 90nm chips.
What happens when you try to run them faster?

A good read on IBM/AMD 90nm SOI CPUs

Source: Overclockers.com

Comments

30 Nov 2004 ~ 10:43pm deepsea The article equates clock speed with processor capability. Not explicitly, but the implication is that AMD is quiet because the processors are low end. Uninformed, or perhaps mislead by the marketing.
30 Nov 2004 ~ 10:47pm Thrax AMD is quiet because users aren't willing to take a flying leap onto the 90nm AMD bandwagon when the only 90nm chips available from them are in the budget range. That's the implication. It's fairly accurate.
30 Nov 2004 ~ 11:34pm mmonnin

Why AMD Promoting its 90nm CPus More?!?

Someone make that a sentense, it doesnt make sense at all. And Capitalize U in CPU.
30 Nov 2004 ~ 11:35pm Thrax "OMGWTFBBQ WHY ARENT AMD MAEK cPu LOLO?!"
1 Dec 2004 ~ 12:30am shwaip

mmonnin wrote:

Someone make that a sentense, it doesnt make sense at all. And Capitalize U in CPU.

Please tell me that your assorted errors in grammar and spelling were intentional.
1 Dec 2004 ~ 1:03am muddocktor Personally, I don't know why AMD isn't selling higher end 90 nm A64 procs, from what I've seen out of mine so far. The damn thing runs so much cooler than my P4's it's not even funny. I've been messing with my 90 nm socket 939 3000+ and my Epox 9NDA3+ for a couple of weeks now and I'm slowly getting pretty impressed with it's performance and also the performance of the Epox board too. I've been slowly ramping up the overclock on it, but running some fairly loose ram timings to minimize ram problems. Once I find what the proc's max stable configuration is I will be working on tightening up the ram timings a bit and then I'll decide if I want to swap this machine out with my Pee4 rig as my main machine. This is where I'm presently at with it and I'm still cooling it with the stock retail hsf, since I've misplaced my A64 hardware for a SLK948-U that I plan to use for cooling. The temps shown in this pic are with the machine folding a Gro wu, so it's loaded to 100%:
1 Dec 2004 ~ 1:15pm edcentric Too bad OC has been breathing too many fumes comming off of their Prescotts.
They have it backwards. AMD isn't making slow chips. I believe that they knew full well of the voltage (and hence speed) limitations of SIO. It is Intel that has screwed the pooch. By making the P4 pipelines so long they enabled/required (depends on point of view) very high clock speeds. AMD chose a design that would yield similar performance at lower speeds an dmuch lower power consumptions.
If AMD is being too quiet I'll bet that it has to do with production capacity, yields, or some other such issues.
1 Dec 2004 ~ 2:58pm Omega65

edcentric wrote:

If AMD is being too quiet I'll bet that it has to do with production capacity, yields, or some other such issues.

Intel vs AMD issues aside.....

I for one was wondering exactly why AMD could make 2.4ghz (& limited numbers of 2.6ghz) Athlon 64s but hadn't released any 90nm CPU at these speeds.

If you read all the way through it looks like 1H 2005 we'll see 2.4ghz & higher Athlon 64s with SSE3 instructions added
1 Dec 2004 ~ 6:38pm TheLostSwede Their top end .13 cpu's are selling like hot cakes even if they cost $800+. Why ditch that production if they make money on it? The other problem i can think of is on die cache. The .09 cpu's are pretty cool right now, but at 2.6 and with 1024 of cache, things are a little different. Until a manufacturer starts creating a mobile chipset with dual channel for A64, we won't see any mobile .09 S939 in the near future either.
1 Dec 2004 ~ 9:00pm Geeky1 IMO, AMD's 0.09 micron CPUs have the same problems that IBM's and Intel's do. They may not be as apparent for whatever reason (although I don't know that that's the case) but they're using the same basic technology as IBM, so it's logical to conclude that they'd encounter many of the same problems...
1 Dec 2004 ~ 9:55pm Omega65

Mackanz wrote:

Their top end .13 cpu's are selling like hot cakes even if they cost $800+. Why ditch that production if they make money on it?

You can prioduce more 90nm CPUs than you can 130nm CPUs on a given Wafer Size. If AMD could make then run at that speed they would, to do otherwise indicates problems.

Remember it took AMD a whole year to move from the original 800mhz sample Clawhammer CPUs to the 1.4-1.8ghz Opteron range they introduced in Apr 2003
2 Dec 2004 ~ 5:05pm TheLostSwede I agree on that Omega. But i think they can't produce that many FX chips from one wafer that qualifies, the rest probably goes to DTR/Mobile bins.

Another thought is that they may only have one fab that can produce .09 cpus?
I think their FX testing is rigouros (spelling?) as a lot of the 3500 winchesters i see do well over 2.6 ghz on the standard heatsink. I'm guessing the next cpu will be a 3700 Winchester at 2.4 very soon. The fact that half multipliers isn't supported is giving AMD a headache i think. A lot more cpu's could have been released.
2 Dec 2004 ~ 7:28pm Omega65

Mackanz wrote:

Another thought is that they may only have one fab that can produce .09 cpus?

Good Point. The new Fab space come online late 2005

And it did take 18 months for AMD to go from 1.8ghz to 2.6ghz with its 130nm process. It will probably take a similar amount of time to perfect the 90nm process too.

Next year will be exciting for Athlon 64 CPUs