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: Petz who wrote (75514)	10/14/1999 5:09:00 PM
From: Tenchusatsu	Respond to of 1572935

Petz, you said, "A more flexible 256K may be almost as good as a 256K + 256K L2." I think you're getting a little confused here. Athlon's L2 cache is unified, not divided into instructions and data. But that wasn't your point. A 256K L1 cache (i.e. 128K data + 128K instruction) is certainly an interesting alternative to an on-die L2 cache. I think the pros and cons between that and an on-die 512K L2 cache will have to be carefully weighed. <BTW, according to one of the CuMine benchmarks out there, the CuMine is still just 32K L1.> Actually, this was already revealed at MPF. A larger L1 cache would have been nice for Cumine, but Intel decided that modifying the L2 cache was easier than increasing the size of the L1 cache. And L2 cache improvements, in my opinion, were long overdue. Besides, if the L1 cache size were increased to 64K, then Cumine-128 wouldn't make sense, unless the core was redone once again to allow for a 32K L1 cache. ;-/ I don't know where the web sites like Firingsquad and Sharky Extreme got the idea that Intel was going to increase the L1 cache size. Tenchusatsu