NVIDIA's GeForce 7950 GX2 & Forceware Rel. 90
PureVideo Updates and Performance
For our next round of tests we took another look at Digital Video processing performance between the two competing core GPU architectures, "PureVideo" technology at work for NVIDIA and "AVIVO" driving ATI.
To characterize CPU utilization when playing back WMV HD content, we used the Performance Monitor built into Windows XP. Using the data provided by Windows Performance Monitor, we created a log file that sampled the percent of CPU utilization every second, while playing back the 1080p version of the "The Living Sea" video available for download on Microsoft's WMVHD site. The CPU utilization data was then imported into Excel to create the graph below. The graph shows the CPU utilization for a GeForce 7950 GX2 and a Radeon X1900 XTX using Windows Media Player 10, patched using the DXVA updates posted on Microsoft's web site (Updates Available Here).
Average CPU Utilization (Athlon 64 FX-60 @ 2.6GHz x 2)
|GeForce 7950 GX2||Radeon X1900 XTX|
The GeForce 7950 GX2 and Radeon X1900 XTX were roughly on par with one another, CPU utilization-wise, when playing High-Def content in Windows Media Player 10. The 7950 GX2 put up slightly lower utilization numbers versus the Radeon X1900 XTX, but the difference was less than 2%. We should also note that depending on the particular video being played, CPU utilization can change to favor either architecture.
Next up, we have a relatively new addition to the HotHardware testing arsenal, the HQV DVD video benchmark from Silicon Optics. HQV is comprised of a sampling of SD video clips and test patterns that have been specifically designed to evaluate a variety of interlaced video signal processing tasks, including decoding, de-interlacing, motion correction, noise reduction, film cadence detection, and detail enhancement. As each clip is played, the viewer is required to "score" the image based on a predetermined set of criteria. The numbers listed below are the sum of the scores for each section. We played the HQV DVD using the latest version of Intervideo's WinDVD 7 Platinum Suite, with hardware acceleration and PureVideo extensions enabled.
|* Updated June 7, 2006|
As we mentioned in our GeForce 7900 GTX evaluation a few months back, both NVIDIA and ATI have been working on new drivers that would drastically improve their performance in the HQV video benchmark. It seems that NVIDIA has come on very strong with their PureVideo support in the new Forceware Rel. 90 drivers, and ATI continues to improve their score in this "benchmark" with each driver release as well. Both companies have improved their score over time thanks to enhanced noise reduction, image sharpening, and inverse telecine algorithms.
We've given both architectures the same score of 113, but must provide some explanation because ATI and NVIDIA do not produce the same results. We initially gave a score of 108 to ATI here, but after scrutinizing the results in the Detail test, it was clear ATI deserved a higher score. ATI does not give users the ability to tweak their edge enhancement algorithm via the Catalyst Control Panel, and their technique doesn't produce as dramatic a change as NVIDIA's in its current state, but there is a definite improvement in image quality in this particular test.
Our Noise Reduction results are also somewhat deceptive. In the Noise Reduction tests, ATI does a very good job of eliminating noise from the various scenes, and some scenes are near perfect. NVIDIA does a good job as well, but depending on what level is set in the driver, NVIDIA's technique could produce some noticeable artifacts. So, even though we give both architectures a 5 in these two tests, ATI currently does a better job. We hesitate to give either architecture a 10 in these tests because whenever there is a color transition, some noise is still clearly visible.
This confusion is ultimately the result of how a score should be determined in HQV. To score a perfect 10 in the Detail and Noise Reduction tests, little to no noise must be visible in the various scenes. To score a 5 there must be some improvement, and to score 0 the scene must remain unchanged. According to HQV's scoring guidelines, we would not give either architecture a perfect score of 10, but they both improve the scenes, which is why they both scored 5s. There is a bit of subjectivity in these tests though, please keep that in mind. We tend to be conservative with our scoring in HQV, but in a "best case scenario", ATI would score 123 here.
Another thing we should bring to your attention is that NVIDIA no longer requires a separate purchase to take full advantage of PureVideo. Whereas users were previously required to purchase a copy of NVIDIA's PureVideo decoder to enjoy all of the benefits of PureVideo, now many popular third-party players like WinDVD and PowerDVD can take full advantage of PureVideo's features. Great move NVIDIA.