Directory  Reviews Graphics Card Reviews |
| Cable-less Crossfire - P965 and the X1600XT |
|
|
|
Page 4 of 5 Our test platform consisted of the following hardware...
Tests were performed under the following OS environment
We kept the test suite to a minimum owing to time frame constraints, utilizing the following software.
All tests were run 3 times, with the best run used for comparison purposes. Our intention here is to evaluate the scalability of a cableless Crossfire configuration, and do so specifically with the P965 chipset, rather than an all out evaluation of the x1600xt itself. We pitted it against our venerable (if loud as a leaf blower) Sapphire x850xt, as tested in our previous review of the Gigabyte GA-965P-S3 which you can read here. First up, 3DMark 2005. Being a benchmark coded for Shader 2.0 supporting cards, we expected the x850xt to have a significant advantage over the x1600xt in overall 3DMark performance, and this was borne out in the tests. With Crossfire enabled, performance of the x1600xt's jumped significantly, roughly 33% faster than a single card. 
Next, 3DMark 2006. Since this is a benchmark which relies heavily on Shader Model 3.0, we expected to see the x1600xt at a slight advantage here, and again this is borne out in the tests with the x1600xt having about a 12% advantage in 3DMark over the x850xt. In Crossfire mode, the twin x1600xt's outperform a single card by nearly 37%. 
Finally, we take a look at Valve's Half Life 2, chocked to the gills with insane headcrab head-humping action. Here we recorded a 60 second timedemo in the "Follow Freeman" section of the game, experiencing much in the way of zombie love and head humpers galore. We used Fraps 2.7.4 to record framerates through the 60 second run at two different game settings. Setting one consisted of a run at 640x480 with all rendering features dumbed down to bare minimum to attempt to take the graphics card out of the equation. Setting two consisted of running at 1280x1024, with all settings cranked to the maximum, with antialiasing set to 6x, and ansiotropic filtering set to 16x. In the first run at 640x480, we take a look at a single card comparison between the x1600xt and the x850xt. As you can see with the higher overall fillrate of the x850xt it stays consistently ahead of the single x1600xt. 
In the second run at 1280x1024, again we're looking at single card performance, and as is borne out by the graph, the x850 edges even further ahead overall. Still, the single x1600xt still offers acceptable performance in this venerable game. 
The third run is telling. With Crossfire enabled, the twin x1600xt's edge out ahead of the single x850xt in overall performance, though not to the levels indicated in the synthetic benchmark runs in 3DMark 2005 and 3DMark 2006. The x1600xt's in Crossfire stay roughly 10% ahead of the last generation x850xt. 
Overall, we'd have to say this is the level of performance enhancement we expected with these cards in Crossfire mode. Not only that, but this has personally borne out for us that the P965 chipset and it's somewhat funky x16/x4 slot arrangement doesn't detract appreciably or noticeably from Crossfire scalability in comparison to the 975 chipset. |
|||||||