RSS Feed

HCW Tech Blog

For the latest info on computer hardware, tech, news, video games, software tips, and Linux, check out our new improved front page: HCW Tech Blog

Reviewed by: Trevor Flynn [11.12.03]
Edited by: Carl Nelson
Manufactured by: Abit
Price: ~$90

Discuss this article in the forum!
Registration NOT Required!


The Performance

As this is my first Intel motherboard review, I unfortunately do not have any competing boards to stack the IS7 up against. However, the main reason a lot of people will be looking at the I865 from Abit is because of it's "BIOS Optimizations" *cough* PAT *cough* support. To see just how much added peformance is gained from these optimizations, I've tested the motherboard at all four different settings: Auto, Turbo, Street Race, and F1.

For testing purposes the IS7 was setup with P4 3.0 800mhz processor, 512MB of Corsair XMS 3500 DDR RAM (Set to SPD), a Soltek SL-GF4TI4200 (41.09), a Maxtor 30GB ATA133, 7200rpm, 2MB HDD, and a clean install of Windows XP SP1. The latest Intel Chipset Drivers were used.

On to the benches!

We'll start things off with some memory benchmarks. I noticed that when I changed the BIOS setting from Auto to Turbo that the latency settings on the RAM were dropped from 2.5-3-3-8 to 2-3-3-6 so there should be a nice little jump at that point. The latency settings were constant from then on out so any increases at higher optimizations are strictly due to internal settings.

First up as always is Sandra 2003

The big jump from Auto to Turbo can no doubt be accredited to the drop in timings. However there are still some minimal increases at the two higher levels.

Aida32 memory tests are next.

Same basic trend here with big jumps from Auto to Turbo and smaller ones after. Notice how increases in Write times don't really jump after the latency drops. This would point to some sort of PAT'like implementation as the prefetch technology increases memory read and not write speeds.

We'd expect the trend to continue through the gaming benchmarks, but just to be safe, let's check them out anyway. Let's start the ever-popular synthetic 3dmark series of benchmarks.

In 3dMark 2001SE, the performance ramps dramatically with the various tweak settings. In 03, it is more limited by the video card used in the test. Even with that limitation, there is still a difference in performance.

Next let's check out another one of Futuremark's synthetic tests with PC Mark 2002, which gives individual scores for the CPU, Hard Drive and Memory

We're seeing the exact same results as with the prior tests. CPU scores are identical, as they should be here.

Next Page: (5)