Kula, Geil Memory question.
登录 | 论坛导航 -> 华新鲜事 -> 技术の宅 | 本帖共有 35 楼,当前显示第 31 楼 : 从楼主开始阅读 : 本帖树形列表 : 返回上一页
作者:Kula (等级:8 - 融会贯通,发帖:22073) 发表:2004-09-22 12:26:37  31楼 
yes, we already save S$300+ coz we got a 3700+(sandra's PR)I'm also curious about those articles... but I think reason lies in many aspects: 1. they got a low-class CPUs or they choose to use MB which is not good at OC 2. our information are out-dated, they post their results around May or April, that time K8T800Pro/Nforce250GB MB for 754 are not popular and are expensive. 3.later they succeeded but many of them didnt make it public 4. we are lucky enough :) I believe with sempron 3000+, the 754 A64 platform will become mainstream for AMD users, higher one is 939 at the end of 2005, or mid of 2006, we shall get a Dual Core K8 with 90nm for acceptable price, and bythen 64-bit platform is ready (OS, driver, application...) or we get K9 directly :D as Microsoft's DirectX plan, mid 2006 or early 2007 DX 10 will come out, with their new OS, longhorn. we change platform then :)
so Vlink 8x is the cause?
I have an idea, let's put the nice blue head sink on SB from NB, and get a fan to activly cool the NB (there are 2 holes around NB for the fan)

AGP 8x is directly link to memory, have anything to do with the Vlink?
Vlink is to transform all system IOs' data from SB to NB, and According to VIA, the Vlink have a fix freq at 266Mhz. If fsb increase, Vlink should remain same otherwise ABit must do something wrong. Still no idea how it crush the agp.

maybe we can suggest Abit to lock the Vlink speed. If fsb get higher, it doesnt have impact on system.
Ping MM: Request Time Out
Ping Money: Destination Unreachable
Ping Life: Unkown Host
Ping Future: Reply From Hell: 0bytes time=0ms TTL=1
欢迎来到华新中文网,踊跃发帖是支持我们的最好方法!原文 / 传统版 / WAP版只看此人从这里展开收起列表

本帖共有 35 楼,当前显示第 31 楼,本文还有 N-1 层楼,要不你试试看:点击此处阅读更多 >>



请登录后回复:帐号   密码