Viper MOTD: Doing Taxes Sucks.
 



















Extreme Computer Mod
Cooler Master Aero 7
OCZ PC3500 DC Kit
Corsair TwinX PC3200
Vantec 470W PSU
Vantec CopperX HSF
2 x 75CFM or not?
Crucial 6 Card Reader
MSI CR52-A2 CDRW
Chaintech AV-515M




 
 
Abit AT7 Max
 
 
Date: September 4, 2002
Manufacturer: Supplied by
Written By:

Testing

I'm going to be testing this motherboard as a direct replacement for my old Asus A7V266 with the VIA KT266 chipset. I'll be testing the AT7 MAX with both PC2100 and PC2700 ram. Each benchmark will be run 3 times and the average scores taken to ensure consistency between results. Main thing I will be looking for here is to see what kind of increase the new motherboard will provide over the older KT266 chipset, as well as of course it's overclocking prowess. The usual line-up of benchmarks will be used here, so let's see what this board can do.

Test Systems

1.4 Athlon Thunderbird

A7V266/AT7 Max

256 meg of Crucial PC2100 Ram

256 meg of Kingmax PC2700 Ram

GeForce 4 Ti4600 (standard clock settings)

Via 4in1's 1438's

Detonator 29.80's

Windows XP

As I stated earlier I am going to test both PC2100 and PC2700 in the AT7 Max to see what if any diffrence the board alone makes to the scores. During the PC2100 testing, the FSB was at 133. Now when I swapped the Ram for the PC2700, there was only a marginal difference in the scores. So why is this? Well, adding PC2700 is not enough on it's own. The Ram and the Northbridge of the motherboard may be communicating at 166mhz, but the CPU having an FSB of 133 is not, so the extra potential of the communication of 166mhz between the Ram and the Northbridge is wasted. To see the true difference that the KT333 chipset can make you have to make sure that the FSB, Northbridge and the Ram are all communicating at the same speed. As the AT7 Max has the 1/5 Divider this is a simple thing to do. Setting the FSB of the CPU to 166 and lowering the multiplier to 8.5 gives us the same speed (or close to) as before of 1400mhz, and having the divider at 5:2:1 means that the AGP and the PCI are still running at there standard 66mhz and 33mhz respectively. I also had to raise the DDR voltage from 2.55v to 2.65v to get full stability from the board.

SiSoft Sandra

First test here is the CPU Arithmetic benchmark.

In this instance you can see that the scores are all of a muchness, with the AT7 max coming in just a little bit ahead. The difference is very marginal indeed.

The Multi-Media benchmark shows a steady increase in performance with the two AT7 Max results being ahead of the A7V266 by a clear amount. Utilizing the 166 FSB on the CPU provides an increase in the performance here as is to be expected.

Previous Page - The Bios

Next page - Testing (continued)

 
     
 
 

Copyright © 2001-2002 Viper Lair. All Rights Reserved. Site Design by