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: Kevin K. Spurway who wrote (57052)	5/3/1999 12:06:00 AM
From: Cirruslvr	Read Replies (1) \| Respond to of 1573895

K7 question JC (new location - jc-news.com has found out that Intel will eventually go back to sockets eventually with Coppermine. I was thinking, I read somewhere (I think it was The Register, but I don't how much to trust them) that AMD may also go to a socket with the K7. If AMD would want to put on-chip cache on the K7 eventually, it would have to be at least 512K (according to Scumbria's 4x L1 cache rule). Since that would make the whole K7 about 140mm2 on the .18 process (according to PB's estimate siliconinvestor.com, the the "K7-2" would be pretty big. Since the K7's bus is such that the memory and cpu have something like separate "buses", do any of you think it would be possible to put L2 cache on the motherboard and run it above the rest of the system's 200MHz? Or is 200MHz the limit until 400MHz becomes a reality next year or whenever? Or could up to 800MHz RDRAM be used as L2 cache? I supply the questions, y'all supply the answers. ;)

To: Kevin K. Spurway who wrote (57052)	5/3/1999 12:58:00 AM
From: Tenchusatsu	Read Replies (1) \| Respond to of 1573895

<Show me some Winstone 97 benchmarks.> Ah, I see. So are you hoping that the K7 will be much faster than a Pentium III on Winstone 2000 than Winstone 99? But, to be fair, the K7 isn't only supposed to increase the performance per MHz, despite the 9-issue execution unit, the 128K oversized L1 cache, the deeper buffers, the 200 MHz point-to-point bus, yada yada. As Scumbria correctly points out, the K7 is designed for higher frequencies than the K6 could ever achieve, just like the P6 core is designed for higher frequencies than the Pentium. So will AMD be debuting the K7 at a clock speed faster than 500 MHz? Seems like they'd have to; otherwise, the K7 will be very underwhelming. That's why AMD hasn't even announced a starting clock speed yet; they're desperately trying to crank out enough parts which can exceed 500 MHz by the time June rolls around. And as Scumbria points out, June is only four weeks away. Tenchusatsu