Did Intel rig test or out engineer AMD/ATI?

When Humphrey Cheung at TG Daily wrote a glowing review of Intel's G965 embedded graphics chipsetbeating an AMD/ATI X1600 dedicated graphics board on a video playback quality test, AMD cried foul, and other journalists echoed some concerns with the massively one-sided test. That prompted Cheung to write a follow-up article, "Did Intel rig its integrated graphics chip demo against AMD?" The test in question is the HQV benchmark and more specifically, the interlacing test, which shows a US flag blowing in the wind--the worst-case scenario for interlacing artifacts.

Interlacing is a grotesque artifact of video that was shown using an every-other-line approach, typically with 60 fields (30 frames) a second. In that test, Intel passed in spectacular fashion using special optimizations in Cyberlink's ClearVideo acceleration and smoothing routines that were optimized for the Intel G965, while the AMD/ATI X1600 failed miserably with the built-in DVD playback software included with Windows Vista. This raised some serious suspicions, and Cheung wrote:

But we aren't the only ones who were puzzled at the side-by-side comparison because Intel showed the same demo to other journalists, journalists who have told us that they will also run their own tests. Scott Wasson, editor in chief of The Tech Report, emailed us saying that his lab guys are testing out the card and will publish the results in a future article. Another editor, Charlie Demerjian from the Inquirer, said that the demo looked strange and told us that they may do their own testing. As a side note, Charlie told us that he offered to bring an AMD to the demo to help Intel correctly configure the computers, but Intel, as you can imagine, declined.

While Intel representatives insist that everything was configured correctly, AMD officials say the opposite. Will Willis, AMD's senior public relations manager, told TG Daily, "The X1600 is unquestionably a much better card for video playback." He added that it's easy to cherry-pick tests that make one product look worse than the other. "We want to let the world know that this one test is not the whole story," Willis said.

To prove his point, AMD cranked up its internal testing labs and didn't use just the flag demo, but all the tests in the HQV benchmark suite. Unlike the Intel demo, though, it played back video from the built-in player/DVD decoder of the 32-bit version of Windows Vista instead of the Cyberlink player. According to Willis, the G965 was thoroughly trounced, in HQV benchmark scores, by the X1600. He also stressed that the scores came from drivers that are currently available as compared to Intel's upcoming driver.

So it sounds like we have a smoking gun against Intel right? Well not quite, and I'll explain why. Not a single person in the above quote addressed the issue of severe interlacing head on, though Scott Wasson said he was going to conduct some tests. AMD basically ran a bunch of benchmarks and said that it received a much higher score than Intel, which is to be expected when you're benchmarking a discrete graphics adapter in the $120 range against embedded Intel graphics costing around $30. But nowhere did AMD's Will Willis address the issue of video playback quality and the grotesque interlacing problem other than to call this "cherry-picking tests". The problem is that video quality is a key criterion for the enjoyment of DVD movies, and it isn't cherry-picking to point out severe Interlacing problems. So the real question is: Is there an interlacing problem or not?

Some of you might say that I'm treating AMD with too much of a critical eye. I'm going to say not so, because I'm privy to information that the journalists quoted above aren't. Some of you may recall when I posted my Long list of bugs that delayed my adoption of Windows Vista two months ago; one of my complaints was about interlacing problems in DVD playback. In fact, the problem was universal in ATI, NVIDIA, and Intel (in Vista's default DVD player/codec) for some DVD titles, and I even managed to get confirmation from Microsoft that this is a known issue. Here's what a Microsoft spokesperson told me regarding the interlacing problem:

The DVD playback issue you’re experiencing is a known issue and Microsoft is actively investigating it.

Windows Vista is actually the first version of Windows that comes with DVD playback capability and it works fine for the most part.  Its primary weakness is how it handles interlacing problems in some DVD titles, which were mostly widescreen video-source anamorphic titles in my experience and confirmed by Microsoft. While I didn't use an HQV interlacing test, it would likely cause major artifact problems, which is what Humphrey Cheung witnessed in Intel's demonstration. Was this caused by deliberate sabotage on Intel's part? My personal experiences with playback problems across all video card platforms, including ATI, and Microsoft's confirmation seems to fly in the face of that possibility.

Intel got around this problem by working closely with one of the dominant DVD playback software companies, Cyberlink, and to optimize Cyberlink DVD playback software with Intel hardware, which is just good engineering. The fact that AMD/ATI didn't do so is AMD's problem--unless Cyberlink went out of its way to cripple those optimizations for ATI hardware, and Intel is implicated in a conspiracy with Cyberlink. If AMD believes it has a case against Intel and Cyberlink, it can take it before the courts like it did for Skype. Last year, when Skype admitted to me that it had programmatically blocked lower-end Intel and all AMD chips from using 10-way conferencing, I sounded the alarm to my colleague Russell Shaw, and he ran with the story. That case was a slippery slope because it's one thing to not optimize for a certain piece of hardware but it's another to go out of your way to programmatically cripple hardware to make it appear that one brand is more capable than the other. Now I'm not saying Intel was in a conspiracy with Skype or with Cyberlink, but the Skype case seemed to be a bit shady, and the courts will need to work that one out.  As for this case, and barring any new court challenges from AMD, this seems to be a case of better engineering rather than a rigged demo.

I'm going to follow up with Intel, AMD, and Microsoft to get to the bottom of this problem, and I've requested lab demos from Intel and AMD in their Silicon Valley campuses.  Ultimately, the user doesn't care whose fault the problem is; they just want their DVDs to look good.

[poll id=19]

.