AMD 90nm processors unimpressive?

[Mephistopheles]

According to <A HREF="http://babelfish.altavista.com/babelfish/trurl_pagecontent?lp=fr_en&trurl=http://www.x86-secret.com/?option=newsd&nid=795" target="_new">some rumors</A> and early 90nm-A64 reviews, the 90nm A64s don't do any better than 130nm and don't quite scale as expected. It's babelfished, but here it is:

In short, the process 90 Nm of AMD, as it is currently, does not seem to be the awaited Messiah who should allow a strong rise in frequency. The very weak frequency of Opteron Dual Core presented at the IDF also seems to go in this direction. Let us hope however that the future revisions will improve this established fact.

So apparently, AMD itself is not without its problems, even if it is facing these problems at a much, much better relative position to Intel.

 

[Spitfire_x86]

I think it's currently like T-bred "A" Athlon XPs. Despite consuming less power, they didn't scale well at all, compared to Palomino Athlon XPs. But later T-bred "B" fixed all scaling issues

------------
<A HREF="http://geocities.com/spitfire_x86" target="_new">My Website</A>

<A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new">My Rig</A> & <A HREF="http://geocities.com/spitfire_x86/benchmark.html" target="_new">3DMark score</A>

 

[Mephistopheles]

In any case, it's slightly disappointing that the 90nm transition wasn't like the 130nm one. Northwood was indeed like 5% or maybe even 10% faster than willamette, like promised.

Prescott, on the other hand, wasn't. And Intel still said it would be like 10% faster than northwood, which made us all be very excited about it, but it turned out to be unimpressive at best.

AMD also tried to pull the "90nm transition is good" thing by saying 90nm athlons would have +5% performance gains, but as it turns out, the new cores are <i>not</i> better than the old ones at all. Again.

 

[trooper11]

i highly question that article and if thats your only evidence, you really dont have much.

first of all, check out whats going on in the overclocking forums. notice the 2.6-2.75ghz speeds on AIR that users are hitting, thats certainly no dissappointment, in fact thats what many were hoping for.

now lets remember that when there was the proposed 5% increase, that was with sse3 additions. well that didnt happen for this revision, so now it looks like the E revision will feature that magical 5% increase you want.

but really, i dont care about that 5%, thats really somehting i could take or elave. the transition to the new process is whats important to me, and it seems that evne in this first revision, amd has succeeded. whats with all the doom and gloom. is it since intel had problems that any even rumors that amd chips arent looking so good is embraced and taken as gold?

i dont know, but i see no real evidence the 90nm transition was dissappointing, and im basing that on user impressions

 

[Madsn]

If i remember correctly the shift from 18nm to 13nm was not only a straight shrink i also think copper was added to the 13nm process

 

[EugeneMc]

<A HREF="http://www.madshrimps.be/?action=getarticle&articID=230" target="_new"> Athlon64 3200+ 0.09µ Socket 939
“The Next 2.4C?”</A>

It seems the opposite to disappointing. Good overclock, runs cooler and with less power consumption (the same conclusion that in Tech Report article) and in benchmarks tested, it's faster than 130nm at the same frecuency clock.

<P ID="edit"><FONT SIZE=-1><EM>Edited by eugeneMC on 10/11/04 08:39 PM.</EM></FONT></P>

 

[wolverinero79]

I thought there was copper added to 180nm processes.

I'm just your average habitual smiler =D

 

[Xeon]

Wow pretty gay site... madshrimps that surely sounds like something related to computers and that we can take it seriously.

Xeon

<font color=red>Post created with being a dickhead in mind.</font color=red>
<font color=white>For all emotional and slanderous statements contact THG for all law suits.</font color=white>

 

[Crashman]

Yes, I remember waiting FOREVER to get a T-bred, then they finally released the crappy T-Bred A when the P4 was scaling like hotcakes and overclocking well. I gave up and ordered a P4 system for overclocking...only to have the order fall through, and bought a P4 2.4B with 512MB PC1066 on an i850E board...for $150. Remember how long ago that was? Then everyone told me I should have saved money with AMD

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>

 

[trooper11]

dont know what gripe you have wiht madshrimps, but ive been following thier articles for a while and i have sene nothing wrong with them. Dont know what reason you have to insult them.

 

[Mephistopheles]

Some <A HREF="http://www.xbitlabs.com/news/cpu/display/20041011124532.html" target="_new">extra info</A> posted on Xbitlabs actually puts the credibility of anyone saying that A64s at 90nm are actually faster in doubt.

AMD only did a die shrink.

No new performance tweaks, no new features... Maybe they're saving those up for later? Or maybe they just wanted 90nm ASAP... who knows?... I know I for one can't complain about their current lineup. It's quite strong.

<P ID="edit"><FONT SIZE=-1><EM>Edited by Mephistopheles on 10/11/04 09:43 PM.</EM></FONT></P>

 

[Xeon]

One word... "madshrimps".

Xeon

<font color=red>Post created with being a dickhead in mind.</font color=red>
<font color=white>For all emotional and slanderous statements contact THG for all law suits.</font color=white>

 

[EugeneMc]

?

 

[trooper11]

....yet 90nm has been a success in my eyes, dont you think so?

besides, amd never said youd see a performance increase clock for clock.

 

[endyen]

Amd has had a successful shrink. No core changes, and no huge power usage. Seems good to me. I just wish Intel had taken that route. Anyone who thinks that woody on 90 nanos wouldn't be shipping now, at 4 gigs is ....

 

[Mephistopheles]

Well, <A HREF="http://www.xbitlabs.com/news/cpu/display/20040715155708.html" target="_new">there were rumors</A> about 90nm offering additional performance...

But things are looking pretty good for AMD anyway.

 

[trooper11]

oh we all knew about the RUMORS, but thats all they were.

im just saying you cant be dissappointed when rumors dont materialize, there was no reason to trust them at all...

 

[endyen]

What seems really unimpressive is the intel 925px (ok so it's xe) chipset. I had hoped the faster speed of the clock would have reduced the effect of the latencies.