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 (55621)	4/15/1999 5:07:00 PM
From: Scumbria	Read Replies (1) \| Respond to of 1577195

Ten, On the other hand, I feel the 200 MHz bus will take more advantage of RDRAM than the 133 MHz bus of the Camino chipset. This could be a real advantage for the K7, if it can get proper chipset support for RDRAM. I still haven't figured out which desktop applications are memory bandwidth limited. I know that Winstone is not one of them, so I don't see K7, DRDRAM or Camino offering any advantage Winstone advantage. Scumbria

To: Tenchusatsu who wrote (55621)	4/15/1999 5:18:00 PM
From: Gopher Broke	Read Replies (1) \| Respond to of 1577195

Isn't the company with the fastest processor going to command the highest margins? Getting ahead of the race early is important for AMD. Does it really matter if we have to wait for the faster bus? In fact, upgrading the bus later gives a nice MB performance ramp-up to go alongside the CPU speed ramp. So the key question is whether AMD can deliver on their promise, not whether K7 will sell with "only" a 133 MHz bus.

To: Tenchusatsu who wrote (55621)	4/15/1999 5:28:00 PM
From: Ali Chen	Read Replies (1) \| Respond to of 1577195

Ten, <No, they'd rather wow you with high "potential" bus speeds, when the reality is that they aren't needed...> The reality is that you are slipping down, boy, in your desperation to bash AMD. In the reality there are more than one address space in PC. There are AGP space, PCI space, I/O space. All these spaces may compete for CPU access, and faster CPU bus helps a lot. Just an example, a CPU may want to read data from memory (1/2 bus bandwidth), process them (in pipeline), and paste the data into AGP space (another 1/2 bandwidth) for displaying. There could also be a snooping activity due to busmastering PCI transfers, for sound mixing for example. I am curious, what are you doing in "server chipsets" if you do not understand these elementary things? Just to remind you, all the data streams are usually well FIFO buffered/combined in the North Bridges (chipset in your terminology), and therefore all transactions can go as efficient bursts across the CPU-to-bridge connection, with good bus utilization. So, calm down, life is good. - Ali

To: Tenchusatsu who wrote (55621)	4/15/1999 6:32:00 PM
From: Kevin K. Spurway	Read Replies (1) \| Respond to of 1577195

Re: "Then the processor-to-chipset connection doesn't need to run at 200 MHz at all. Run that bus at 133 MHz and the K7 will do just fine." Other stuff interfaces directly to the chipset as well (PCI, AGP, etc.). This consumes some bandwidth I'd imagine, especially with 4x AGP, where the "extra" 66 Mhz could come in handy. Plus, marketing a 200 MHz FSB is better than marketing a 133 or a 100 MHz FSB, even if technically the 200 MHz FSB is only between the CPU and chipset. <GG> Kevin