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: Tenchusatsu who wrote (40506)	10/31/1998 1:53:00 PM
From: RDM	Read Replies (1) \| Respond to of 1573682

When evaluating on-chip cache the speed difference between the cache fill rate and the cpu rate is of critical importance. The on chip L2 cache of the K6-3 will help less at 450 Mhz than it will help at 550 Mhz. In fact, there is diminishing performance benefit to increasing the clock rate of a K6-2 after a point due to the restrictions imposed by 100 Mhz bus.

To: Tenchusatsu who wrote (40506)	10/31/1998 1:55:00 PM
From: Scumbria	Respond to of 1573682

Tench, Sure, that L2 cache becomes an L3 cache, but the L3 cache isn't likely to make much of a difference due to diminished returns. Winstone does show a significant performance difference between a 256K and 1MB cache. K6-3 systems will likely perform better than Dixon systems at the same clock speed. Scumbria

To: Tenchusatsu who wrote (40506)	10/31/1998 1:56:00 PM
From: Maxwell	Read Replies (1) \| Respond to of 1573682

Tenchusatsu: <<Face it, Jim, adding 256K of on-die L2 cache is not going to improve the K6-3 as much as an on-die L2 cache improved the Mendocino Celeron, since the K6 and K6-2 already had an L2 cache on the motherboard. Sure, that L2 cache becomes an L3 cache, but the L3 cache isn't likely to make much of a difference due to diminished returns. Plus, there is no reason for me to believe that the "core changes" within the K6-3 are going to improve its per-clock performance. More likely, those core changes were necessary to get to the higher frequencies in the first place.>> New core includes "write combines" which increases about 5% over old core. K6-3 has been benchmarked. It is about 15-20% faster than K6-2 and outperforms PII clock for clock. Maxwell

To: Tenchusatsu who wrote (40506)	10/31/1998 6:20:00 PM
From: Jim McMannis	Read Replies (1) \| Respond to of 1573682

Ten, RE:"Face it, Jim, adding 256K of on-die L2 cache is not going to improve the K6-3 as much as an on-die L2 cache improved the Mendocino Celeron, since the K6 and K6-2 already had an L2 cache on the motherboard. Sure, that L2 cache becomes an L3 cache, but the L3 cache isn't likely to make much of a difference due to diminished returns. Plus, there is no reason for me to believe that the "core changes" within the K6-3 are going to improve its per-clock performance. More likely, those core changes were necessary to get to the higher frequencies in the first place." ------ I see my friend Maxwell has already posted information for your enjoyment. I'll yield to him since he's very sharp about these things. As far as my record, which and what benchmark you use will have some bearing but I'll venture a guess that the K6-3 will outrun the current Pentium II easily....Intel will catch up if it incorporates 256k cache on the Celeron. Jim