NVIDIA's GeForce 7950 GX2 & Forceware Rel. 90

HOW WE CONFIGURED THE TEST SYSTEMS: We tested the NVIDIA based cards on an Asus A8N32-SLI nForce 4 SLIX16 chipset based motherboard. The ATI powered cards, however, were tested on an A8R32-MVP motherboard based on the CrossFire Xpress 3200 chipset. Both systems used the same AMD Athlon 64 FX-60 dual-core processor and 2GB of low-latency Corsair XMS RAM, though. The first thing we did when configuring these test systems was enter each BIOS and loaded their "High Performance Defaults."  The hard drive was then formatted, and Windows XP Professional with SP2 was installed. When the installation was complete, we installed the latest chipset drivers available, installed all of the other drivers necessary for the rest of our components, and removed Windows Messenger from the system.  Auto-Updating and System Restore were also disabled, the hard drive was defragmented, and a 1024MB permanent page file was created on the same partition as the Windows installation. Lastly, we set Windows XP's Visual Effects to "best performance," installed all of the benchmarking software, and ran the tests.

The HotHardware Test Systems AMD Athlon 64 FX Powered

GeForce 7950 GX2 Performance Comparisons with 3DMark06 v1.0.2 Details: http://www.futuremark.com/products/3dmark06/

3DMark06

Futuremark recently launched a brand-new version of their popular benchmark, 3DMark06. The new version of the benchmark is updated in a number of ways, and now includes not only Shader Model 2.0 tests, but Shader Model 3.0 and HDR tests as well. Some of the assets from 3DMark05 have been re-used, but the scenes are now rendered with much more geometric detail and the shader complexity is vastly increased as well. Max shader length in 3DMark05 was 96 instructions, while 3DMark06 ups the number of instructions to 512. 3DMark06 also employs much more lighting, and there is extensive use of soft shadows. With 3DMark06, Futuremark has also updated how the final score is tabulated. In this latest version of the benchmark, SM 2.0 and HDR / SM3.0 tests are weighted and the CPU score is factored into the final tally as well.

To make the graphs in this article a bit easier to read, we've grouped the dual-card and single-card configurations into two sections.  The top of the graphs are all technically single-card configurations, although one could argue that the 7950 is two cards. The bottom of the graphs are made up of dual-card configurations. Also note that we've tested the GeForce 7950 GX2 at both NVIDIA's reference clock speeds and XFX's shipping clock speeds, which are significantly higher.

As you can see, the new GeForce 7950 GX2 is one heck of a performer, regardless of whether or not it's clocked at NVIDIA's reference specs in XFX's "pre-overclocked" state. The single 7900 GTX and Radeon 1900 XTX cards are simply left in the dust according to 3DMark06.  The SLI and CrossFire rigs fare much better, as well they should considering they cost hundreds of dollars more than a single 7950 GX2.

If we break down the results into their individual components, we see that all of the NVIDIA powered cards perform very well in 3DMark06's shader model 2.0 tests, edging out their ATI powered counterparts by a fair amount.

In the HDR and shader model 3.0 tests, however, ATI's Radeon X1900 XTX is a fair bit faster than the 7900 GTX. The 7950 GX2's two GPU configuration is untouchable in this test though without running a high-end SLI or CrossFire rig.  Overall, the new 7950 GX2 is roughly 40% faster than a single 7900 GTX in 3DMark06 and it trails a 7900 GTX SLI setup by about 17% depending on its GPU and memory clock speeds.