Author: Jeremiah Penery
Date: 09:09:58 05/08/02
Go up one level in this thread
On May 08, 2002 at 09:36:22, Aaron Gordon wrote: >On May 07, 2002 at 21:37:37, Jeremiah Penery wrote: > >>On May 07, 2002 at 12:16:30, Aaron Gordon wrote: >> >>>Also, about running 166/200fsb synchronous. Like I said.. the 8k3a+ and other >>>similar boards have 1/5, 1/6 PCI multipliers and similar AGP multipliers to keep >>>everything within spec while keeping the bus and ram at 166 or 200MHz(333 & >> >>I was talking about running the CPU bus at the normal 133MHz and running the >>memory at 166MHz (333 DDR) - _a_synchronous operation - it usually doesn't >>produce much speedup, because the processor bus (still at 133MHz) is already >>saturated, no matter if the memory bus is putting out 20GB/s. > >It's not always saturated. You'd need a RAID array blasting away with a highend >Geforce4 in 4x agp mode to get it nice and bogged down. Argh. No, no, no! I'm NOT talking about the PCI bus or the AGP bus or anything else. In chipsets, there is a bus from the northbridge (memory controller) to the main memory, and a bus from the northbridge to the CPU (as well as a bus to the southbridge and to the AGP controller). If you run ASYNCHRONOUSLY the memory bus will be different speed than the CPU bus. The CPU bus has a maximum theoretical bandwidth, depending on the processor bus speed you select - raising the memory bus speed DOES NOT AFFECT THIS! If you run FSB at 133, this bandwidth is 2.1GB/s, even if you run the memory at 200MHz (400 DDR). If you are getting any more than this from the memory, the CPU bus will be saturated already, and performance will receive only the tiniest increases. Which is why running with a huge memory speed doesn't help unless the actual FSB is increased to match.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.