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: Dan3 who wrote (108775)	5/1/2000 3:28:00 AM
From: Tenchusatsu	Read Replies (3) \| Respond to of 1571863

Dan, <I guess I still have one question: is it cheaper or easier to manufacture or design a quad pumped 100MHZ circuit than a double pumped 200MHZ circuit? Or is it basically just two ways of doing the same thing?> No, there are significant differences between quad-pumping a 100 MHz bus and double-pumping a 200 MHz bus. Remember that double-pumping and quad-pumping only affects the data bus. Everything else runs at the speed of the main clock, such as request signals, cache snoop signals, bus arbitration, etc. Also, whenever there is a dead cycle inserted in-between two data streams (e.g. one driver finishes its transfer and gives up control of the data bus to the next driver), that dead cycle must be one main clock cycle long. That means a dead cycle on a quad-pumped bus would represent four data transfers lost, while the same dead cycle on a double-pumped bus would represent only two data transfers lost. As a result, a quad-pumped bus is less efficient than a double-pumped bus. In short, it's much easier to move from a double-pumped data bus to a quad-pumped one than it is to double the frequency of the main clock. But in terms of performance, it's better to do the latter than the former. Tenchusatsu