ATI vs. Nvidia in a Video Shootout

ThinSkin writes "ATI and Nvidia are well known for hailing their products as leaders in 3D apps and games, but little is known that both companies are trying to stake their claim in the video market as well. ExtremeTech is featuring an article that tests cards from ATI and Nvidia to determine who takes the cake in video quality and performance. Using CPU utilization scores and visual quality comparisons during video and DVD clips, the author concludes that ATI's latest generation of GPUs have an edge over Nvidia, particularly in DVD playback and with video acceleration."

Comments

New algorithm

[2.7182]

I thought that Nvidia had the edge because they are using the new fast subdivision algorithms of Jean Gallier at Penn CS dept.

Re:New algorithm

[Feminist-Mom]

I heard that too on a CAGD newsgroup, but I think it is still in negotiation. Gallier's work is interesting though - here is his webpage , which has all of his articles.

Re:New algorithm

[Anonymous Coward]

"2D "quality" is now largely unimportant." - by RzUpAnmsCwrds (262647) on Tuesday January 31, @03:55PM (#14610450)

Well, considering I spend a good 90% of my time on PC's in 2d Explorer Windows shell display (even though I like to game, I work coding during the day & spend MOST of my time web-surfing for technical info. here @ night to try to stay ontop of the change in this field) @ both home & work?

I consider 2d display important, & especially vs. eyestrain & from what I understand, the higher the refresh rate? The better it is for your eyes.

(Please - DO feel free to correct me here, or if there is 'upper-limits/returns-on-investment' as far as hz refresh rates in regard to eyestrain prevention/vision health - thanks! I can stand to learn as much as the next guy, & tend to pick up something new, everyday, from diff. folks with diff. outlooks on various portions of this field).

"If you want the best quality display, you use a good LCD display with DVI - and the Silicon Image TDMS transcoders that NVIDIA cards use are just fine." -

Personally? I use Trinitron CRT monitors, & for the reasons I state above - they tend to yield very fast refresh rates, are proven technology (I have heard, for instance, that LCD Flatpanel types STILL have the "dead/stuck pixel" problem on them to this day quite a bit).

Plus?

I can 'snag' a USED Trinitron for around $45-55 U.S. Dollars around where I live for a 17", & $75-$85 for a 19", & know they last (I have been using the latter size for 3 years now & bought it used, but, I may have gotten lucky too - it's possible!)

However, I have seen Trinitron 17" units (4 nights ago in fact) running off a GeForce Ti4600 push 200mhz refresh (but, I am NOT sure what color range he was using, it may have been less than "True Color/32-bit" etc.) + his using the 40.72 drivers on Windows XP - that's FAST, & for a pal of mine who was considering going to flatpanel LCD types because he thought his monitor was slow!

He had it set to 60hz only for his refresh rate in Display Properties, Advanced, Monitor section!

Anyhow - I have found that buying used, has helped me out, as far as monitors are concerned (costs especially & the amazing longevity I have gotten out of Trinitron units as well as their speed of display (and the 2 tiny lines don't bug me personally, but they may others)).

APK

P.S.=> An 'example given', from my experience with other CRT type monitor vendors over time (others' experiences may vary, but this was one of mine in buying new... I do NOT think I ever will again, & here is why):

A 21" ViewSonic Graphics Series I had for instance that I paid $850 for in 1999?

Well, it went sour the first year out!

So, I paid $80 U.S. Dollars to ship it to ViewSonic for a replacement (they shipped back the upgrade to it no less, nice, but read on) & again, w/in exactly almost 1 year, it too went sour!

So, in my experience with CRT type monitors @ least - Trinitrons are faster that those by other OEM's (refresh-rates being so high) & just last longer! :)

* Some "food-for-thought" on purchasing Trinitrons, & also used monitors (especially of the trinitron variety because of their longevity (@ least in my experience & anecdotal example/proof to you))... Going "used" CAN be a good buy, & save you TONS of cash! apk

Re:New algorithm

[aaronl]

Everything since the GeForce FX series cards can do 2048 × 1536 at 85Hz
The GeForce 2 could do 2048 x 1536 at 75Hz
The RIVA TNT could do 1600 x 1200 at 85Hz

Your NVIDIA board has dual 400MHz RAMDACs, and that ATI card had dual 400MHz RAMDACs, so they have the same sync capabilities. If you can't push higher than you are, it's because your *monitor* can't sync at that frequency. Many monitors won't do 1600x1200 at over 75Hz.

It also wasn't urban legend about ATI's drivers being terrible. They still have issues today, though they is *much* better than before. ATI still loves to strand people with prior generation cards, though. Their control software is also one of the worst ones out there, courtesy of whatever bright star decided to write it in .NET, and poorly at that.

Re:New algorithm

[SillyNickName4me]

First of all, I just love tinitron displays for the picture they produce, esp. for gaming (black level and variable refresh with vsync), video (black level, variable refresh with vsync, 'analog scaling' and depending on your display, native interlaced video) and graphics work where it is important that colors on your screen get as close to 'real' as possible. I have a small collection of them from 15" upto 30"

For heavy text editing and the like however I really do prefer a large TFT screen.
Why? all the things a trinitron CRT is good at and typical consumer TFT screens not won't matter for text editing at all while a TFT screen has some real advantages for it:

What you say about refresh rate and eye strain is true, but somewhat specific to CRT and similar technology, but doesn't apply to TFT screens.

Because there is no beam tracing lines on a screen, all the pixels are being displayed at the same time, there is some time involved in pixels having to change color, but the traditional flickering of a CRT simply does not occur at all, resulting in a stable and flicker free picture.

Then, a TFT screen doesn't have to deal with a whole bunch of inherently analog electronics to control and aim a beam of electrons, While a TFT screen is quite complex to make, it is surprisingly simple in its operation, and there is a lot less that can go wrong with again the result of a more stable picture.

Add DVI to the things above and you end up with a result that is pretty much impossible to match with a traditional CRT with regards to again getting a stable picture.

If you have to work with text a lot, do yourself the favor of trying one for a while if you can.

Re:New algorithm

[Anonymous Coward]

"First of all, I just love tinitron displays for the picture they produce, esp. for gaming (black level and variable refresh with vsync), video (black level, variable refresh with vsync, 'analog scaling' and depending on your display, native interlaced video) and graphics work where it is important that colors on your screen get as close to 'real' as possible" - by SillyNickName4me (760022) on Tuesday January 31, @05:56PM

Yup, same areas I love them for (entertainment in DvD or other film format playback & gaming performance)... plus, price (if you buy used that is as I often do because of having been 'burnt' once imo, buying a 21" new which I mention last posting).

* Of course, I could have just gotten 'lucky' in the Trinitron I bought that has outlasted ones I paid literally 10x much for new, vs. this trinitron unit used.

"For heavy text editing and the like however I really do prefer a large TFT screen. Why? all the things a trinitron CRT is good at and typical consumer TFT screens not won't matter for text editing at all while a TFT screen has some real advantages for it:" - by SillyNickName4me (760022) on Tuesday January 31, @05:56PM

Well, I am "ALL EARS", especially regarding this technology, so I will read-on & comment as I go, quoting ya!

(I can stand to learn about this technology since I admittedly have NOT used one for long periods first hand @ home or @ work etc.)

"What you say about refresh rate and eye strain is true, but somewhat specific to CRT and similar technology, but doesn't apply to TFT screens.

Because there is no beam tracing lines on a screen, all the pixels are being displayed at the same time, there is some time involved in pixels having to change color, but the traditional flickering of a CRT simply does not occur at all, resulting in a stable and flicker free picture." - by SillyNickName4me (760022) on Tuesday January 31, @05:56PM

Good points about "TFT" monitors (assuming this is modern flatpanel types, which I know next to nothing about, other than 'hearsay' online) & ones I was not aware of either!

However, the ONLY 'sticking point' I have with them is what I have heard about "dead pixels" (sticking ones?) that tends to happen with them still/nowadays even on 'state-of-the-art' ones.

(I haven't used flatpanel/tft monitors myself in the workplace or home over the years, so I am really "out-of-it" as to what they offer in the way of better eye-health etc. (well, supposedly I have heard they send out less radiation as well vs. CRT types, a GOOD 'health-related' point in their favor & they DO take up less desktop space, a definite "+" in their favor imo)).

"If you have to work with text a lot, do yourself the favor of trying one for a while if you can." - by SillyNickName4me (760022) on Tuesday January 31, @05:56PM

I do, extensively, for work (writing code) & home (surfing is my main fun @ home, studying up for work in fact, or this field in general really)...

I will check them out sooner or later, & I tend to be a late-adopter/late-bloomer & 'set in my ways' (hard for me to change until I take a try @ new things + I do USUALLY wait until 'bugs shake out' & buy new technologies a few generations AFTER their release because of this (this lcd/flatpanel/tft stuff is not "new" anymore though, so it may be time to give it a shot when finances permit, just as I did with ATI cards a year or two ago FOR A YEAR vs. my fav. NVidia))!

Anyways - your post was enlightening to me regarding this technology for display since I admittedly know VERY LITTLE about it!

APK

P.S.=> Likewise - I hope mine was as informative to others in like regard, for those not aware of some of the points I made @ least (probably doubtful, because Slashdot does have (usually) some pretty 'enlightened folks' hanging about it for the most part from this field & many others from my readings here)... nice talking with/to you! apk

Re:ATI cards are good...

[harrkev]

What I want to know is if ATI still wins under Linux. It is really cool to say that ATI has the best video playback, but if you are building a MythTV box, a test under Windows does not really tell you much.

video editing the new war

[Brigadier]

At first I thought big deal but then it accured to me that of all the people I know ( your typical family pc) the most common use is to download and edit pictures, and video. I am amazed how quickly a pc newbie user can become a proficient video editor with just a few tools. I'm sure it wont be long before they double or tripple the pc gaming market share. will be nice one day to see the prices of DV cards come down with the main streaming of things like firewaire and digital video for the common home user.

Linux Driver Reviews??

[gasmonso]

Many PVRS support linux and the number increases every year. Since this article deals with DVD/DIVX movies and not gaming, I would like to see some reviews with Linux drivers. Anyone have any experience?

http://religiousfreaks.com/

Re:Linux Driver Reviews??

[laffer1]

If you're building a PVR, stick to a generic tv tuner card. ATI rage theater chipsets are not usually supported (AIW, etc) in most open source operating systems. Something more generic will work almost anywhere (even *BSD).

In general, nvidia cards are better in open source operating systems in part because nvidia actually writes drivers for linux, freebsd, and solaris to some degree (now oss). I love ati cards, but my love of BSD trumps that. I do have to say the fx 5200 card i bought looks great on the console and runs enemy territory at a playable level in freebsd 6 with xorg 6.8.2 and linux emulation. I tried my aiw 9600xt in linux a few months ago and it looked sweet. i can't give a framerate comparison because the nvidia card is in a different pc and one of them is a dual xeon and the other is a sempron.

Video on Linux

[Anonymous Coward]

What is the state of video on Linux?

I would love to see a comparison of performance and video quality of these same cards on Linux. Do the drivers even support any of this functionality? Is CPU usage similar?

Honestly

[asv108]

I don't even look at ATI anymore when building a system for my own use. Nvidia has had excelent Linux device driver support for a number of years now. The last few personal systems I built were nvidia dualhead systems running linux, and I have never had a driver problem.

My latest system is dualhead dual-dvi pci-express 7800GT system running on Ubuntu. I was expecting the video configuration to be a major pain the ass, but everything worked well.

Until ATI has the same level of Linux support, I will not take their products under consideration.

Re:Honestly

[apoc.famine]

I've had a Radeon 9700 for a few years...the first 4-5 months it sat on my shelf, because the WINDOWS drivers were too poor for me to bother using it. At that time, I had far better performance from my Geforce3 Ti5 card. After 4-5 months had passed (and the price had dropped $50) I was finally able to use it, as they had improved the drivers. Still, it had far more issues than my Nvidia card had.

Fast-forward a year or so, and when I was drawn into the dark side and installed linux, the ATI drivers were appallingly poor. My old Geforce3/XP1900 system outperformed my Radeon 9700/XP2800 system by a decent amount in most games. In addition, I could never get updated ATI drivers to work without first using their utility to reconfigure my xorg.conf, then TOTALLY rebooting the machine...just restarting X didn't do it. And with some issues getting direct rendering working, it took a *lot* of xorg.conf reconfigures and reboots.

After using the ATI card for about six months, I finally gave up using it and got a Geforce4 6600 GT card, and haven't looked back. Quick, easy install, drivers work decently well, and I don't have to reboot/reconfigure every time I update them. Restarting X works just fine. And hearing more and more stories about ATI's linux drivers I'm in the same boat - they wouldn't be on my list when building a linux system. But after my experience with the windows drivers, they aren't on my list at all.

Three out of four systems that I've owned in the last 5 years or so had Nvidia cards and worked well. The one which started with an ATI card was a pain in the ass until I got rid of that card. While it could be a bad card, a lot of the people I know with ATI cards haven't been overly impressed. Until Nvidia starts dumping out crap, I guess I'm sticking with them.

Video capture? AVIVO?

[MrBandersnatch]

..but little is known that both companies are trying to stake their claim in the video market as well.

And both are going to fail prity miserably while they fail to provide serious technical information on their video capabilities. I've a need for H.264 *encoding* accelleration and video capture atm but trying to get information on the exact capabilities of cards (especially AVIVO) was a PITA. Sometimes the marketing droids would e far better beingg replaced by a technician.

Anyways, pity the article doesnt look at anyhing apart from DVD playback - to be honest, how high CPU utilisation is while playing back a DVD is a long way down my list of priorities when Im looking at buying upto 8 £400+ cards. How about capture quality, driver stability etc etc?

Not Microsoft's fault

[MobyDisk]

If you want your video to look its best and run as fast as it can, you have to enable all sorts of settings in the advanced properties of your player (or players, plural), and those settings can be different between ATI and Nvidia cards.

Yuck.

In short, Microsoft needs to seriously clean up this mess. Video codecs need to hook into a common framework, one that the graphics cards manufacturers can target for acceleration without needing to work with every individual codec maker on the planet.

This is an interesting statement, because the author just described exactly how DirectVideo works. Each step in the decoding process is a pipeline, and a "codec" can plug-in to this and provide whatever steps in the process that it can do best. For example, if playing a video looks like this:

Read a DVD -> Reading a file -> Decrypting -> Decompressing -> Motion compensation -> YUV2RGB -> Deinterlacing -> Scaling -> Displaying on video device -> ATI X1800

There can be a separate component registered for each step. Or many. And DirectVideo can determine which one is the most appropriate for the given input, output, and hardware configuration. So if you video card supports hardware YUV2RGB scaling, then it will do it. If not, the software can.

The problem is partially that crappy companies get in the way. I downloaded a codec so I could view DV files, and it registered such that all video types were DV. This is a common scenario that requires a purely brain-dead programmer:

boolean IsThisTheProperCodecForThisVideoType?(string videoType)
{
// TODO: Look at type code and see if it is a DV file
return true;
}

Re:Remember When

[Eightyford]

Back in the day you could judge the quality of a video card by how fast it displayed the "stars" screensaver on windows 3.1 .. And the truly awesome rigs wouldn't skip every few seconds.

And you could also use the solitaire falling cards test. It actually used to take minutes for all the cards to fall after winning a game of solitaire.

who cares about video encoding?

[Tibor the Hun]

Buying a windows machine for video encoding and DVD authoring is like buying a Mac for games.

ATI Linux

[DaCool42]

For those of you wondering about linux drivers - ATI's fglrx linux driver works fairly well (I use it to play HD .ts files on a Radeon 9800 pro). The only problems are lack of support for xvmc, and some problems with dual head (confusing config, xinerama issues). I don't have any performance issues with full bandwidth 1080i content and 5.1 sound running on a 720p display (video de-interlaced with mplayer's halfpack filter).

Nvidia wins

[WhiteWolf666]

I run Linux.

ATI has gotten better on Linux, but Nvidia vastly outperforms ATI on Linux.

I would not recommend anyone purchase an ATI card for Linux usage, and I wouldn't commit to maintaining anyone's system if they have an ATI card.

For 2D, or Video, they are okay, but they are severly lacking for OpenGL usage.

VLC versus Elecard for HDTV

[dunc78]

I downloaded VLC to play HDTV video clips that I downloaded from my cable box via firewire and I was not able to get it to play the clips without dropping a signficiant amount of frames; however, Elecard MPEG Player was able to play the clips without dropping frames. So it seems as though VLCs codecs are not as efficient as others. The details are 720P/60FPS video on a Athlon64 3200, 512 MB RAM (Single Channel), with an AGP-8x PNY GeForce6600 (256 MB). I may not have had something configured right on VLC, but I fumbled around for hours trying to get it working.

Re:VLC versus Elecard for HDTV

[Xesdeeni]

Yeah. Comparing 2-D playback of DVDs in 2006, is like comparing 3-D frame rates using Quake II...passe.

Riddle me this Batman:

1. Can the card accelerate MPEG-2 playback (DxVA, et al)?
1.a. How much CPU is necessary to play back HD content (720@24p, 720@60p, 1080@24p, 1080@30i) without dropping frames?

2. Can the card accelerate MPEG-4 (h.264 part 10) playback?
2.a. How much CPU is necessary to play back HD content (720@24p, 720@60p, 1080@24p, 1080@30i) without dropping frames?

3. Can the card accelerate WMV (VC-1) playback?
3.a. How much CPU is necessary to play back HD content (720@24p, 720@60p, 1080@24p, 1080@30i) without dropping frames?

4. Can the card accelerate MPEG-2 encode?
4.a. How much CPU is required to get real-time encode (i.e. 1 hour of video takes 1 hour to encode)?

5. Can the card accelerate MPEG-4 (h.264 part 10) encode?
5.a. How much CPU is required to get real-time encode (i.e. 1 hour of video takes 1 hour to encode)?

6. Can the card accelerate WMV (VC-1) encode?
6.a. How much CPU is required to get real-time encode (i.e. 1 hour of video takes 1 hour to encode)?

7. Can the card synchronize 1080i video with 1080i display (i.e. the field synchronization between the decoded video and played video don't drift - hint, neither ATI nor nVidia can do this today)?

Xesdeeni

A counter point

[lakeland]

While Nvidia's closed-source drivers are clearly better than ATI's, the opposite is true of the open-source drivers. If you are looking to build a system without binary drivers, or are using non-x86 and so cannot use the provided drivers, then you're better off going with ATI.

I imagine this is no coincidence, how many people can be bothered working on the nv driver when the nvidia driver works so well... But it does worry me how easily we have come to accept binary drivers now that they work so reliably for 90% of the users.

Blame Microsoft?!

[kidjan]

Note: I do not represent the opinions of MSFT, nor do I speak on their behalf. The below is my opinion.

The author of the article concludes with this ridiculous statement:

In short, Microsoft needs to seriously clean up this mess. Video codecs need to hook into a common framework, one that the graphics cards manufacturers can target for acceleration without needing to work with every individual codec maker on the planet.

A few observations, as someone who has done extensive programmatic work for digital video in windows:

1. Video codecs in windows do hook into a common framework--that framework is called DirectShow. It's specifically designed to process audio and video, and includes highly advanced features that until several years ago were not even present in *nix (see GStreamer for the *nix equivalent to DirectShow), and many features that currently are not (to the best of my knowledge) available in *nix. This framework is completely free, has a substantial portion of open source code (see the baseclasses), is used in almost every media application MSFT makes and the majority of many custom players, and has been available for in excess of a decade. So, anybody not using it is A) stupid and/or B) technically inept.
2. This framework *does* allow acceleration, to a limited degree, but the problem has nothing to do with Microsoft. In order to provide acceleration, Microsoft has to work with NVidia and ATI (which I assure you, they are) closely--it cannot be a singular endeavor. I had to remove DXVA from the product I'm currently working on because the drivers being provided by NVidia and ATI were too unstable for us to realistically release the program into the wild. How and why should I "blame Microsoft" for that?

By no means is Microsoft saintly or innocent (far from it), but it seems to me that they just can't win no matter how they play the game. The statement above is just looking for a quick target rather than addressing the real problem: people who are too dumb to make codecs that leverage a standards based playback architecture (it doesn't even have to be DirectShow--there are other architectures out there). DirectShow is a very developed, very extensive framework for processing audio and video, and it is solely the fault of people proliferating the market with excessive, buggy, redundant code that there are conflicting third-party applications.

Were MSFT to do anything to "fix" this problem, they'd have to further restrict restrict codecs in DirectShow, in which case the above author would proceed to whine about how MSFT doesn't allow third parties enough integration. Having your cake and eating it too? I think so.

What about Matrox?

[Xabraxas]

2D quality on Matrox cards is outstanding. How come we couldn't get a comparison with on of their cards. I have a Parhelia laying around here somewhere but unfortunately it's not quite working anymore (the screen is a nice shade of pink).