trentcys

1. Intel聲稱Netburst嘅branch prediction係業界最勁原來唔係吹水
詳情請睇:
http://www.hkepc.com/bbs/viewthr ... &extra=page%3D5


2. Conroe 2MB & 4MB L2行SuperPi 嘅分別冇以前想像嘅大...用PrefMonitor搵到答案~
下圖係Northwood行SuperPi 32MB嘅L2 hit rate...

行32M每過幾秒會hit rate會短暫下跌, 但512K/1M就冇. 原因係就算運算資料大過cache容量, CPU/compiler識得將data預先放入L2, 令L2 hit rate可以相對保持高位.
Conroe嘅道理都係一樣, 就算只有2MB都可以靠prefetch data黎保持L2 hit rate.
(估唔到data prefetch嘅準確度都相當高下~)

Richteralan

The accuracy of this tool is questionable.

I\'m getting the same result as you with my Opteron 144 @ 2.25GHz

BeleBala

引用:

Originally posted by Richteralan at 4/8/2006 08:50:
The accuracy of this tool is questionable.

I\'m getting the same result as you with my Opteron 144 @ 2.25GHz

果排棒棒顯示最高記錄唔係 96.2% ga @@

我粒 Prescott 2M 99.4% =.=

qcmadness

引用:

Originally posted by BeleBala at 2006-8-4 10:18:

果排棒棒顯示最高記錄唔係 96.2% ga @@

我粒 Prescott 2M 99.4% =.=

因為Conroe既IPC高

XT

我果陣玩完都 O 鬼晒咀

qcmadness

引用:

Originally posted by trentcys at 2006-8-2 20:55:
1. Intel聲稱Netburst?駼ranch prediction係業界最勁原來唔係吹水
詳情請睇:
http://www.hkepc.com/bbs/viewthr ... &extra=page%3D5


2. Conroe 2MB & 4MB L2行Sup ...

其實佢做到咁勁既branch prediction, 係因為架構問題.......
我諗而家AMD做Northbridge既能力比以前都有大增...