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.

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

To: wanna_bmw who wrote (162734)	3/21/2002 5:27:03 PM
From: dale_laroy	Read Replies (1) \| Respond to of 186894

>Yes. That is the overhead that I was talking about. Logically, adding extra ways of associativity are trivial. All you have to do is design it in. Simple. The trick is managing the trade-offs. If you get 5% by doubling the associativity, but lose 8% due to a longer look-up time, then what have you accomplished?< What you may have accomplished is performance scalability. Perhaps doubling the associativity gains 5% at 1.5 GHz, 6% at 2.0 GHz, 7% at 2.5 GHz, etc. If the loss from longer look-up remains at a constant 8%, or more likely declines along the lines of 8.0% at 1.5 GHz, 7.8% at 2.0 GHz, 7.6% at 2.5 GHz, etc. the performance scalability would favor higher associativity over shorter lookup time as the design is shrunk.