Our Experience with UVD and R600
It's easy to point to the presentations and say, "look, this is slightly misleading." This doesn't get across the even more convincing interactions we had with AMD personnel over the course of reviewing the HD 2900 XT. Hopefully this will help to illuminate the issue further. I'll have to put a disclaimer on this and say that I don't have a photographic memory and I forgot to bring my diary to the press briefing. This isn't normally the kind of stuff I take notes on for product launches.
At the AMD press event in Tunis, Tunisia (we still can't figure out why we went to Africa to talk about a graphics card), myself and a couple other reviewers had the good fortune of being able to see HD Decode first hand. We watched a side by side comparison of an R6xx based part and a G8x based part with interesting results that mostly pointed out the long way Intervideo and CyberLink have to go in refining their player software. We were under the impression that this was R600, though clarification was not provided.
We talked about UVD and PureVideo, and we were asked if a video decode comparison would be made against NVIDIA hardware for our HD 2900 XT launch article. Upon replying in the affirmative, we got the impression that we should expect the R600 to outperform its competition (the 8800 GTS) in video decode tests. This doesn't seem as certain without UVD, but our marketing speak filters failed to pick up on anything out of the ordinary.
Moving on, we were initially told to expect a driver to enable video decode acceleration shortly before launch. AMD also provided us with a copy of PowerDVD which was said to support AVIVO HD. We did get a newer driver near launch, but in testing it we saw very bad video decode performance. It turns out that we couldn't test video decode at all, as hardware acceleration was not enabled in our press driver. This indicates to us that, even though we saw decode acceleration working in Tunis, AMD was having more trouble than expected in the QA department.
We would expect QA issue with new UVD hardware. If we've got the same video decode hardware as a previous generation part, it seems a little odd to think that we wouldn't have a driver or software to enable the feature.
In going back through our slides for this piece, we noticed that it is explicitly stated that UVD is included on the Radeon Mobility HD 2300. This is an R5xx based part retrofitted with UVD. Of all the parts not to have this new feature, the DX9 class Mobility HD 2300 would have been the one we thought of first. Surely if this part has the new feature, the HD 2900 XT should also include it. The argument for not is, again, that higher end GPUs tend to be paired with CPUs that can handle the workload.
The reason we offer time and again for the fact that G86 and G84 include full video decode while G80 does not is that the slower parts had 6 months more development time. In NVIDIA's case, it makes complete sense that G80 would lack the more refined video decode hardware. For AMD to launch a top to bottom product line and leave a feature like this out of their highest end part is very strange.
So strange, in fact, that combined with all the other experiences we've had with this rather surreal product launch we still just can't help but doubt what we're hearing. To us, the most sensible explanation for all this is that the hardware must be there but is somehow broken - perhaps broken to the point that it will never be fixed on current silicon (i.e. via drivers) and a new spin would be required to address the problems.
In spite of what makes sense to us, AMD has indicated that UVD hardware is not at all present on R600. Their "official" statement has yet to appear, but we'll update this article when/if we get it.
It's easy to point to the presentations and say, "look, this is slightly misleading." This doesn't get across the even more convincing interactions we had with AMD personnel over the course of reviewing the HD 2900 XT. Hopefully this will help to illuminate the issue further. I'll have to put a disclaimer on this and say that I don't have a photographic memory and I forgot to bring my diary to the press briefing. This isn't normally the kind of stuff I take notes on for product launches.
At the AMD press event in Tunis, Tunisia (we still can't figure out why we went to Africa to talk about a graphics card), myself and a couple other reviewers had the good fortune of being able to see HD Decode first hand. We watched a side by side comparison of an R6xx based part and a G8x based part with interesting results that mostly pointed out the long way Intervideo and CyberLink have to go in refining their player software. We were under the impression that this was R600, though clarification was not provided.
We talked about UVD and PureVideo, and we were asked if a video decode comparison would be made against NVIDIA hardware for our HD 2900 XT launch article. Upon replying in the affirmative, we got the impression that we should expect the R600 to outperform its competition (the 8800 GTS) in video decode tests. This doesn't seem as certain without UVD, but our marketing speak filters failed to pick up on anything out of the ordinary.
Moving on, we were initially told to expect a driver to enable video decode acceleration shortly before launch. AMD also provided us with a copy of PowerDVD which was said to support AVIVO HD. We did get a newer driver near launch, but in testing it we saw very bad video decode performance. It turns out that we couldn't test video decode at all, as hardware acceleration was not enabled in our press driver. This indicates to us that, even though we saw decode acceleration working in Tunis, AMD was having more trouble than expected in the QA department.
We would expect QA issue with new UVD hardware. If we've got the same video decode hardware as a previous generation part, it seems a little odd to think that we wouldn't have a driver or software to enable the feature.
In going back through our slides for this piece, we noticed that it is explicitly stated that UVD is included on the Radeon Mobility HD 2300. This is an R5xx based part retrofitted with UVD. Of all the parts not to have this new feature, the DX9 class Mobility HD 2300 would have been the one we thought of first. Surely if this part has the new feature, the HD 2900 XT should also include it. The argument for not is, again, that higher end GPUs tend to be paired with CPUs that can handle the workload.
The reason we offer time and again for the fact that G86 and G84 include full video decode while G80 does not is that the slower parts had 6 months more development time. In NVIDIA's case, it makes complete sense that G80 would lack the more refined video decode hardware. For AMD to launch a top to bottom product line and leave a feature like this out of their highest end part is very strange.
So strange, in fact, that combined with all the other experiences we've had with this rather surreal product launch we still just can't help but doubt what we're hearing. To us, the most sensible explanation for all this is that the hardware must be there but is somehow broken - perhaps broken to the point that it will never be fixed on current silicon (i.e. via drivers) and a new spin would be required to address the problems.
In spite of what makes sense to us, AMD has indicated that UVD hardware is not at all present on R600. Their "official" statement has yet to appear, but we'll update this article when/if we get it.
Facebook
Twitter
RSS
53 Comments
View All Comments
trisct - Tuesday, June 5, 2007 - link
The reason behind the UVD confusion seems fairly obvious. ATI was just purchased by a CPU company. The original design for the 2000 series GPUs almost certainly included UVD, it would have been an obvious weakness vs. NVIDIA, even if they only have full decode for one compression standard.The problem is that parent AMD is meddling with ATI's plans, and decided to do an about-face on the UVD component, in order to promote uses for multi-core CPUs. The marketing material was already done at that point, though, so everything got confusing. This is a clear internal conflict of interest, where on one hand AMD wants to sell high-end CPUs, and make them as useful as possible for as many tasks as possible, but also sell high-end GPUs, which in this instance compete directly with one of the main reasons to get a powerful multicore CPU. I hope this all doesn't boil down to ATI graphics going the Intel route and depending on the CPU more and more.
lemonadesoda - Wednesday, June 13, 2007 - link
Probably will.Anyone else get the impression that ATI is suffering major brand dilution. I have been a loyal ATI customer since 8500. My "loyalty" was partly due to driver simplicity (one driver pack for all machines)... but also because you knew-what-you-got and prices and performance was both competitive and leading edge.
I don't hold the same view anymore.
lopri - Tuesday, June 5, 2007 - link
Or take the job offer from NV. He won't be missed.Frumious1 - Tuesday, June 5, 2007 - link
Neither will your posts, if you'd care to STFU and get back to sniffing the AMD happy-shrooms.Khyron320 - Monday, June 4, 2007 - link
People are being way to harsh on this review. It is just good journalism they are reporting the facts here. Now if Nvidia is getting away with something similar and Anandtech is not reporting it then i would call them "anti-AMD". But as far as i know there they have not skipped on any anti-Nvidia news/reviews.lemonadesoda - Wednesday, June 13, 2007 - link
Good journalism = facts + issues + accurate + well writtenI supprt godd journalism. But his article scores a low 50%, and is not up to Mr. Wilsons usual standards. He can do better
DerekWilson - Sunday, July 22, 2007 - link
actually I appreciate the fact that you feel I can do better, and I agree that this was one of my weaker articles ... but I don't think what was said was unimportant or inaccurate ... companies need to be careful how the present details, especially when it can be so easily misconstrued by their partners (let alone journalists).TheOtherRizzo - Monday, June 4, 2007 - link
I think AT are absolutely correct in putting down AMD for this. But the question is why have these kind of things never been investigated before and why have AMD been singled out when Nvidia are just as guilty of lies concerning Purvideo. In case someone from AT is interested here is a quick list of issues to get you started:-Nvidia still [url=http://www.nvidia.com/page/purevideo_support.html">http://www.nvidia.com/page/purevideo_support.html]claims[/url] all sorts of Purevideo features for AGP cards when in fact they have been disabled in the driver for over a year now. Not even a task as basic as hardware deinterlacing works anymore (even my GeForce 3 could do that). This is a far more blatant lie than what AMD did with UVD.
-both AMD/ATI and Nvidia claim H264/VC-1 decoding when in fact they skip deblocking in most cases although deblocking is a non-optional part of the spec.
-both AMD/ATI and Nvidia announced hardware encoding years ago. AMD is still trying to be clever by offering a software only encoder that will only work if certain GPU's are present.
To find more information search doom9 and avsforums.
DerekWilson - Wednesday, June 6, 2007 - link
Thanks for the comments and the email ... Sorry I hadn't stepped in here and updated this thread earlier. I actually haven't been feeling well lately :-(Anyway, I just want you to know that we really do take your comments seriously and I'm looking into the issue. Here's a copy of the email I sent a couple NVIDIA PR and technical marketing contacts:
-----
-----
Hey guys,
In the wake of all the AMD Avivo HD confusion with their HD 2900 XT, I'm getting a lot of requests from users to look into other issues with improper reporting of feature support.
A number of readers have pointed me to this page:
http://www.nvidia.com/page/purevideo_support.html">http://www.nvidia.com/page/purevideo_support.html
These readers are saying that current drivers disable support for most (if not all) purevideo features in all AGP hardware, even with MPEG-2. I haven't tested this yet, and I'm not sure if I'll have the time this week. But I'd really appreciate it if you could get back to me with a statement about this.
If purevideo is currently disabled on AGP hardware, if the feature list is not accurate with current drivers, and if there are any plans to change this, can can you let me know?
I'm sure, with the current issues we're having with AMD, it goes without saying that accuracy and honesty would go a long way here.
Thanks,
Derek Wilson
7oby - Sunday, June 10, 2007 - link
Thanx, but I guess the question is too general to be answered by anyone at nvidia.
I just did a quick google to recall my last findings, but it's not complete nor did I always find the source of information.
WMV AGP Accleration seems to be enabled up to 78.01
http://forums.nvidia.com/lofiversion/index.php?t15...">http://forums.nvidia.com/lofiversion/index.php?t15...
Audio/Video sync issues seem to be the reason for disabling this Accel mode:
http://forums.nvidia.com/index.php?showtopic=12150...">http://forums.nvidia.com/index.php?show...p;mode=t...
There exists an additional theory that the AGP Bandwith is not sufficient for the way the work of the different accelerated decoding stages is distributed among CPU/GPU (German only):
http://www.computerbase.de/forum/showpost.php?p=10...">http://www.computerbase.de/forum/showpost.php?p=10...
About MPEG-2 and SSE related Stuff, I currently don't have any links.
When this topic was of more general interest other reviewers also did not get detailed information (Only German: Read the "update" part):
http://www.computerbase.de/news/hardware/grafikkar...">http://www.computerbase.de/news/hardwar...zember/b...
7oby