Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: Jim McMannis who wrote (57085)	5/3/1999 2:41:00 PM
From: RDM	Read Replies (1) \| Respond to of 1572874

I said in previous posts that 40% gain may possibly obtained by K7 in contrast with current PIII for FPU intensive applications assuming the the K7 is clocked at a 20% higher frequency than the PIII. This was total speculation that a 20% gain for FPU ops on K7 was possible for the same clock rate as the PIII. This was a guess based upon considering the triple pipelined, out of order execution capable, awesome floating point processor in the K7. The current PIII has only two pipes, not three, and does not handle out of order execution well. I believed a 20% clock advantage may possible at some point, not at introduction, for the K7. This advantage may occur as soon as December IMHO.

To: Jim McMannis who wrote (57085)	5/3/1999 2:41:00 PM
From: kash johal	Read Replies (1) \| Respond to of 1572874

Jim, Re: K7 up to 40% faster. I think you are right that folks were comparing the projected K7 against what was available at the time. When the K7 hype started the PII was the processor and clock speeds were 450Mhz with 1/2 speed L2 cache and 100Mhz bus. I suspect that a 600Mhz K7 will be 30-40% faster than that configuration. However in a 4-5 months Intel will be running at 600 Mhz and possibly 700-800 Mhz with 133Mhz bus + RDRAM and full speed cache on chip. Admittedly the next generation Coppermines are 0.18 micron vs 0.25 micron K7's. If AMD can only ship K7's at 0.25 micron then it is not clear to me that it will be any faster than the leading edge Coppermines. In fact I would expect Intel's 0.18 micron to ramp easily to 700+ megahertz considering they can get good yields at 500Mhz with their 0.25 microns. Clearly 0.18 yields and capacity will be the key for AMD. 0.18 will enable a die size much smaller than the current k-3's and clock speeds to > 800Mhz+. It certainly is not clear cut to me AMD can succeed in winning the speed race but it does seem that they should at least be able to match Intel's highest clock speeds. In that event they should certainly jeapordize Intel's high end margins.

To: Jim McMannis who wrote (57085)	5/3/1999 2:47:00 PM
From: Tenchusatsu	Read Replies (3) \| Respond to of 1572874

<I don't recall anyone stating the K7 will be 40% faster.> You ought to read your own links: aceshardware.com That guy says "If AMD can deliver on time (i doubt it), K7 will be at least 40% faster than anything Intel has for more than a year." <So what looked like a possible BIG speed leap ahead of Intel is diminished by the time AMD puts out the K7.> But the way AMD is talking, they make people think that Intel is just sitting in the sun wearing shades and sipping Pina Colladas. Sure, the K7 might have a higher performance limit than the Pentium III. And sure, if everything else were equal, like on-die caches, 0.18 microns, and RDRAM support, the K7 could surpass Pentium III by a sizeable amount. The problem is that everything else isn't equal; Intel will have the lead in on-die caches, 0.18 microns, and RDRAM support, and you can bet Intel will be pushing those advantages along with higher clock speeds in response to the K7. Tenchusatsu