AMD to announce 'HAMMER' Architecture details

<b><A HREF="http://www.amdmb.com/vb/showthread.php?threadid=66311" target="_new"><font color=blue>http://www.amdmb.com/vb/showthread.php?threadid=66311</A></font color=blue>
>>San Jose, CA-October 15, 2001 - AMD's Vice President and Chief Technical Officer Fred Weber today will disclose architectural details of AMD's next-generation "Hammer" microprocessor architecture at Microprocessor Forum in San Jose, Calif. Weber will also discuss how HyperTransport(tm), the innovative technology designed to accelerate internal system communications, simplifies the design and boosts the performance of multiprocessing systems. A question-and-answer session will follow his presentation.

Following his presentation, AMD plans to issue a press release and other information with highlights of the "Hammer" presentation. Information will be available at <A HREF="http://www.amd.com/news/spotlight" target="_new">http://www.amd.com/news/spotlight</A>

Date: Monday, October 15, 2001
Time: 3:10 p.m. U.S. PDT.<<</b>

Good or Bad have no meaning at all, depends on what your point of view is.

"Having a bit of problem -- as I am sure you can guess. I'll try to have it fixed up ASAP! " - HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHaH

I think I hurt myself =P

Second link is as good as the first.

<b><font color=red>????????</b></font color=red>

Good or Bad have no meaning at all, depends on what your point of view is.

Probably not SGI or Sun. But the jury is still out as to whether it's Intel or AMD procs. Not that it really matters.

<font color=green>I post so you don't have to!
9/11 - RIP</font color=green>

Yeah, I was also rather unimpressed.

"Ignorance is bliss, but I tend to get screwed over."

Yes, I had already seen those slides. They did not produce any new information. It would seem all they did was try to increase the spotlight on hammer, not release any actual new information.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =

Excuse me, but have I misunderstood somehting in a big way. Does the hammer appear to have a 32 stage pipe line. This seems to indicate that one instruction takes 12 ns to filter down out of the execution and into memory. I mean the P4 pays through the nose for it s 20 stages. But 32, I must have got something wrong, this can't be right, can it?

Democracy Bernad, it must be stopped!<P ID="edit"><FONT SIZE=-1><EM>Edited by charliec2uk on 10/16/01 06:45 AM.</EM></FONT></P>

Man that presentation doesn't really tell you anything !

I'm wondering if AMD have actually produced any working samples of Hammer yet or at the moment just exists on paper.

I want to see facts, figures ! At the moment Hammer looks like a WONDER processor other than that....not much else.

Here are some rumors:

<font color=blue>In revealing some of the first details about the chip, Weber hinted that it would be released at 2GHz and said that on one key benchmark, SpecINT, Hammer would score nearly twice as high as 64-bit chips on the market today from such as those offered by Intel Corp., Sun Microsystems Inc. and IBM.</font color=blue>
Source:<A HREF="http://www.eweek.com/article/0,3658,s=1884&a=16521,00.asp" target="_new">http://www.eweek.com/article/0,3658,s=1884&a=16521,00.asp</A>

<font color=purple> **Life is too short to be pissed off all the time.**</font color=purple>

so what?
Nvidia made chips for consules that ran MUCH lower res and speed then voodoo's, and the big bang they made once they hit the market with TNT products..
who knows maybe AMD is on to something..

thogh i think amd would release a prospect of the new product at the site if that would be the case, seems like a unsponsered site needed some cash fast..

<font color=green>
*******
*K.I.S.S*
*(k)eep (I)t (S)imple (S)tupid*
*******
</font color=green>

Weird. The slides are dated October 12, 2001, yet they still use the old Virtual address space of 48 instead of 64 and physical address space of 40 instead of 52. I guess the marketing team hasn’t read the white paper.

Schmide

There was one interesting piece of information regarding the memory controller.

Moving the memory controller onto the CPU die is an approach Sun has used on the UltraSparc processors resulting in significant improvements in memory latency and bandwidth.

I don't expect the Hammer to come close to the Ultra performance wise but AMD are obviously willing to follow innovators other than Intel.

<font color=blue> Smoke me a Chip'er ... I'll be back in the Morgan </font color=blue>