Gigabyte GV-R98P 128D Radeon 9800Pro (4)

	<![if lt IE 4]> <SCRIPT SRC="http://imgserv.adbutler.com/adserve/;ID=131075;size=468x60;setID=61435;type=js" LANGUAGE="JAVASCRIPT"> </SCRIPT> <NOSCRIPT> <a href="http://imgserv.adbutler.com/go2/;ID=131075;size=468x60;setID=61435"> <img src="http://imgserv.adbutler.com/adserve.ibs/;ID=131075;size=468x60;setID=61435;type=img" border="0" height="60" width="468"></a> </NOSCRIPT><![endif]>
中文版 \| News \| Archives \| Reviews \| Forum \| $ Deals \| Links \| History \| Contact \| Privacy

Gigabyte GV-R98P128D 128M Review (4)
Overclocker 12 May 2003

Benchmarking

We conduct the a number of tests on the setup with the following configuration :

Test System:

Intel P4 3.0C
ABIT IC7-G i875P running 2 x 256M DDR400 in Sync DDR400 CAS2, 2-2-5
Maxtor D740X-6L ATA133 7200RPM
AOPEN 40X CDROM
Gigabyte GV-R98P128D, Catalyst 3.0
Tyan Tachyon Radeon 9700Pro, Catalyst 3.0
Channelwell 375W AMD Approved PSU

Comanche 4

Comanche 4 is a demanding application and is excellent for testing Direct 3D.

In our benchmarks for Radeon 9800pro, we set it to the 5 Main settings, Optimal Performance, Hi Perofrmance, Balanced, Hi Quality, Optimal Quality. Do take note the various combinations of AA and AF when the above are selected.

Optimal Quality = 4X AA + 16X AF
High Quality = 2X AA + 8X AF
Balanced = Application Specific
High Performance = Application Specific + Texture Pref 3 + MipMap Level 3
Optimal Performance = Application Specific + Texture Pref 2 + MipMap Level 2

In fact, when we used this application to test the cards, there are some strange results. For example, in the Balanced, Hi-Performance and Optimal Performance tests, the R98P128D scores lower than a Radeon 9700Pro. That is quite surprising isn't it. When it runs in AF+AA modes (in Optimal and Hi QUality), we can see a difference between the two cards.

Radeon 9700Pro : 32.6 (Optimal Quality), 44.72 (Hi Quality)
Radeon 9800Pro : 39.07 (Optimal Quality), 50.99 (Hi Quality)

Well that shows that the Radeon 9800pro is still a better performer when you compare it in image quality when it is enabled.

This benchmark also shows one fact that the benchmark might not be suitable for future tests as it is not as demanding in Normal and Performance tests.

Next