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 : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: fyodor_ who wrote (35751)	4/14/2001 1:19:48 PM
From: dale_laroy	Read Replies (1) \| Respond to of 275872

"Also when using the new BIOS (1.09)?" To the best of my knowledge, Ace's is the only site that has used this BIOS. And since Ace's did not test the KT266 with PC1600 DDR SDRAM, the impact of this beta BIOS on PC1600 DDR SDRAM performance is unknown. I would like to see a review that compares the performance of KT266 on the same board (such as the just reviewed Shuttle) using PC133 SDRAM versus PC1600 DDR SDRAM.

To: fyodor_ who wrote (35751)	4/14/2001 1:30:43 PM
From: pgerassi	Respond to of 275872

Dear Fyodor: A quick check on a 64byte L2 cache line load (typical FSB traffic) yields 14 cycles at CAS3 PC133 or 105ns. For CAS2 PC1600 is 9 cycles or 90ns. For CAS2 PC133, 12 cycles or 90ns. But this is at the DRAM bus. Add a cycle of wait for initiation from the FSB for all three cases and we get, 112.5ns for CAS3 PC133, 100ns for CAS2 PC1600, and 97.5ns for CAS2 PC133. Well what do you know, CAS2 PC133 is faster than CAS2 PC1600. If we add one more cycle to pass data back to the FSB from the DRAM bus, CAS3 PC133 gets 120ns, CAS2 PC1600 gets 120ns, and CAS2 PC133 gets 105ns. I think this explains why the DDR PC1600 gets whacked, the shorter initiation and transfer wait states for moving data between buses for 133FSB could overwhelm double the bandwidth at 100FSB. This is also why, super bypass of the AMD 760 allows the CAS2 PC1600 to outrun a waited CAS2 PC133 and its only allowed on the 760 because synchronization is guaranteed. CAS2.5 PC2100 (75ns, 82.5ns, and 90ns) will always have lower latency than CAS2 PC133. Problem solved. Pete