Catalyst 8.8 makes reverse results between HD 4870 and HD 3870?

Resently we posted an article comparing Catalyst 8.8 with 8.7, in which we mainly focused on their behavior on DirectX. Providing us a furthur view, the OpenGL news site Geek3D posted an article titled ‘ATI Catalyst 8.8 vs 8.7: OpenGL Performance Drop’ which is based on HD 3870.

Platform:

CPU: C2D E8400
Motherboard: EVGA 790i Ultra SLI
RAM: 2GB DDR3 1333MHz
VGA Card: HD 3870 @ reference clock
Operation: Windows XP

As the chart shows, HD 3870 face a performance drop on OpenGL when using Catalyst 8.8. All benchmark drop except for Furmark.

However, the increased score on Furmark is also strange. In spite of the HD 3870, HD 4870′s score on furmark actually dropped dramatically(drop by around 50%) in our review. And in our forum, some readers also posted the same issue. We wonder why different cards will achieve different scores on the same driver?

You can read the original article to know more.

2 Responses to “Catalyst 8.8 makes reverse results between HD 4870 and HD 3870?”

  1. dejanh Says:

    Seems that FurMark has become a real thorn in the side of ATI. Because of inadequate VRM cooling on the HD 4870s and the fact FurMark heats up those VRMs a lot FurMark can overheat and even kill most HD 4870 cards. That is the reason behind the drop in FurMark that has been noticed by a lot of people. Basically, instead of fixing the problem with VRM cooling ATI appears to have just nerfed FurMark so that it cannot cause the cards to overheat during the stress test. Some people have suggested that renaming the FurMark executable to something else (e.g., fur.exe) allows to bypass the driver nerf and have FurMark behave the same as it would have in 8.7 or before. GPU-Z 0.2.7 is good for monitoring VRM temps while running FurMark in windowed mode.

  2. レイバン ウェイファーラー Says:

    夫は、自分を律する者は、自分の意志と理性であると考えるようになっていた。神に頼るほど自分は弱くはないと、信夫は自分を考えるようになっていた。なぜなら、あの夜以来、信夫は猛然と自分の情欲と闘って、そのたびに打ち勝ってきたか

Leave a Reply