SI
SI
discoversearch

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.
Technology Stocks : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?


To: Scumbria who wrote (78432)4/11/1999 4:45:00 PM
From: Paul Engel  Respond to of 186894
 
SCUM bria - Re: "A lower latency onboard L2 cache will improve performance, but don't expect to see anywhere near the same kind of benefit from a large onboard L2 that you do with a large onboard L1 (like K7.)"

The higher latency L1 cache of AMD's desKtop 7 will be one of the reasons for the desKtop 7's very, very limited success.

It will be slow and eat up valuable real estate.

At 184 sq. mm., look for Cherry Sanders to use this - and the resultant LOW YIELDS - as an excuse for the AMD losses due in Q399 and Q499.

Paul



To: Scumbria who wrote (78432)4/11/1999 4:47:00 PM
From: Paul Engel  Read Replies (1) | Respond to of 186894
 
SCUM bria - Re: "A lower latency onboard L2 cache will improve performance, but don't expect to see anywhere near the same kind of benefit from a large onboard L2 that you do with a large onboard L1 (like K7.)"

The higher latency L1 cache of AMD's desKtop 7 will be one of the reasons for the desKtop 7's very, very limited success.

It will be slow and eat up valuable real estate.

At 184 sq. mm., look for Cherry Sanders to use this - and the resultant LOW YIELDS - as an excuse for the AMD losses due in Q399 and Q499.

Why don't you tell us how detrimental AMD's desKtop 7 off-chip, 1/3 speed L2 cache is going to be ?

Looks like that slow speed oof-chip cache may cause the desKtop 7 to be a non-starter - don't you think?

Paul



To: Scumbria who wrote (78432)4/11/1999 8:51:00 PM
From: Tenchusatsu  Read Replies (2) | Respond to of 186894
 
<don't expect to see anywhere near the same kind of benefit from a large onboard L2 that you do with a large onboard L1 (like K7.)>

Uh, Scumbria, we've been through this before. You and I both agree that the L1 cache is "oversized" in the case of the K7. And you and I both agree that its longer latency does impact performance somewhat, even if a lot of it is hidden in the pipeline or the out-of-order execution. Of course, the K7 will make up for it with a lower L1 miss rate, which is absolutely vital if their initial K7 comes with off-chip L2 cache which is slower than that of a Pentium II 400.

In other words, Intel will speed up the L2 cache by putting in on the die, while AMD will sacrifice a little L1 cache performance in order to keep from going to the L2 cache in the first place. Different approaches, different requirements, different performance characteristics. The oversized L1 cache is not the panacea that you make it out to be.

Tenchusatsu