Slashdot Mirror
ATI Introduces a Parallel Processing Video Card
bilestoad2 writes " ATI has announced the introduction of a new video card, The Rage Fury Maxx. It uses two RAGE 128 Pro Processors and 64 Megs of Ram.
Follow this link for the complete story.
I don't know about you, but I've got to have one of these..."
Sorry, ATI, but I don't subscribe to your "proprietary technology" anymore. I made the mistake of buying one of your cards before when I didn't know better; I'll not make the same mistake again. There are a lot of cards which are better than yours who actively support free development - like Matrox and nVidia - and I'll not be shackled with your wholly proprietary support - or lack of any support - again.
Well, since at least this doesn't seem to have much to do with the Internet, the Patent Office's "It's Net So It's New" psychosis shouldn't apply.
Of course, they did give 3DFX a patent on Multitexturing, which while more unique than ATI's Full Frame nonsense, is still essentially pretty obvious--and not just after the fact.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
Splitting full frame rendering among multiple processors...this is patent pending?
Are you kidding?
It's at least arguably unique to split even and odd lines among two cards(like Voodoo 2 SLI), or to split the image into horizontal strips(Metabyte's PGC), or to evenly split the texel reprocessing load among multiple texel processors(Voodoo 2 core design), but to attempt to patent the process of merely having one complete frame go to one processor while having the next complete frame go to the next processor?
The general reason one doesn't want to use a full frame architecture is simple: Per frame times don't budge. Either you have to build a higher latency into your rendering chain, since the chipcluster has to know the next x frames you intend to render, or you get *no* speed boost.
Don't even get me started on out of order frame rendering on a realtime rendering solution.
Each of the previously mentioned solutions(SLI/PGC/Texel x 2), incidentally, lowers per-frame latency.
Granted, there's probably some degree of multi-frame latency built into most drivers, particularly for games. But the concept of patenting the most basic parallelization solution strikes me as absolutely hilarious. It's very likely most 3D rendered movies use the technique ATI is trying to patent. "I'm done finishing this frame, send me a new one."
It's very likely most WORKPLACES work the same way too. "I'm done with this job, assign me a new one."
That being said, I'm looking forward to trying out ATI's new cards. Ever since I noticed their 128's were supported by Metabyte's excellent Eyescream system, I've been much more interested in them.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
We've had them for about a month and a half now. The members working on it were delayed (myself, included- had my vacation in that timeframe, ah Scotland!) and we're getting seriously into it at this point. ATI's rep's been fairly helpful.
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
I rag on ATI as much as anyone for being habitually behind the times (and I have used every major video card on the market now except GeForce, which I'm planning on getting fairly soon).
:( ... )
You should know that 3dfx, the recognized performance leader for most of the last 2 years (this of course may be changing) has been using the exact same old technology for quite some time. Voodoo2 is just a someone updated Voodoo1 chip (they added hardware support for the lighting calculation along the edges of the triangle, iirc). Voodoo3 is just an overclocked SLI Voodoo2 (Voodoo3 2000 has exactly the same performance numbers as a V2 SLI rig, as we measured it). Voodoo4 will have some new stuff (tho no one knows exactly what yet), but it is definitely still a direct descendent of SST1, the core Voodoo technology.
Similarly, TNT2 is just a massively overclocked TNT (GeForce does appear to be quite different).
I will give ATI the benafit of the doubt till I see the performance numbers.
Hilariously, the Maxx has 2x as much memory as the high-end Rage 128, for a whopping 64 megs, which is totally absurd. That's as much memory as my entire system! (and of course I can't afford more ram now
You know the weird thing? I just read about Maxx on ATI's webpage yesterday. When I woke up (4 minutes ago) I realized I had a dream about it. Then I came here, and there was a story about it! Man, I must be sad....
Time to go email ATI...maybe I'll ask them for nice open Linux drivers.
Well, I don't know about RC5, but it's possible to implement Conway's Game of Life using the stencil buffer on newer 3D cards (i.e. anything not put out by 3Dfx). There's a wonderful description of it in the back of the OpenGL ARB Redbook. RC5 is probably a bit too complicated to do properly with graphics operations, though...
---
"'Is not a quine' is not a quine" is a quine.
"'Is not a quine' is not a quine" is a quine.
Quine "quine?
I don't know what exactly their patent is trying to cover. It looks like it's trying to cover distributed, rather than parallel, rendering; that is, in triple-buffering, have one chip handle the first backbuffer and the other chip handle the second backbuffer. The law of diminishing returns would definitely apply right away. Right now one of the big bottlenecks in 3D cards is the speed at which the bus can send rendering commands to it. Also, the time it takes to send a rendering command is often longer than the time it takes to execute it on the card; unless each chip is storing a complete displaylist and then post-rendering it (and there's not really much point to that, either), the overlap between the chips' rendering times will be minimal, at best. At the absolute best you could get a doubling in framerate, but the latency would still be just as high, and latency is the real killer in 3D games, not framerate (it's just that framerate is easier to measure and easier to explain).
Perhaps some of their patented work involves trying to 'interpolate' between frames. If that's the case, then that really is a quite difficult problem, and I'd be tempted to say they deserve any patent they get in that area. However, I seriously doubt that's the case.
Basically, this seems like another case of Exxtreme Marketing[tm]. ATI seems to have taken a page out of 3Dfx's book. (I'm sorry, but the Tbuffer is nothing revolutionary - it's a crippled accumulation buffer being marketed as revolutionary, when the TNT and Rage and Savage and the like have had a full accumulation buffer for a couple years now.)
---
"'Is not a quine' is not a quine" is a quine.
"'Is not a quine' is not a quine" is a quine.
Quine "quine?
Okay, so ATI's MAXX is rendering alternate frames on a different chip. Sounds precisely like a 3dfx SLI setup. And if you look outside of the 3D world, there have already been parallel processing display adapters. I once owned a Radius Firestorm 192 which featured three S3 864 chips. Each one was responsible for a color: one chip for red, one for blue, and one for green. The card was really, really fast in its day, and produced the sharpest picture I have ever seen, including that of my current Millenium II.
Anyway, just thought I'd remind ATI of the past.
-jwb
summary:
- nothing new under the sun.
- ATI reinvents pipelining, ignores drawbacks.
Firstly a rant about the press release and its quoted 5% drop between 16 bit frame rates and 32 bit frame rates for this new ATI card: Any manufacturer could do this by artificially limiting their 16 bit fill rate. This number says nothing unless combined with an absolute fill rates at either bit depth.
Now to the deja vu: ATI has effectively shoehorned two cards' worth of acceleration into one graphics subsystem. This has been done twice before in the consumer space: first by 3dfx, with Scan Line Interleave, which allowed two cards to work in parallel on any polygon that spans more than one line on screen; more recently by Metabyte, with their Parallel Graphics Configuration, which partitioned the screen vertically into two independent regions and dedicated a card to handling each of the regions.
Both 3dfx and Metabyte use spatial partitioning to get parallelism. 3dfx could do it finer grained because they had control over the chipset design and could include a mechanism for tight synchronisation of two cards. Metabyte went coarse-grained because they had to do the picture recombination from the two cards in external hardware, and it was hard enough to make this work at all without making it work for alternating scanlines. So why didn't Metabyte save themselves a bunch of hassle and use the "temporal" partitioning (or, in other words, pipelining) approach that ATI is now using? Hmmmm...
One issue here is latency. (For this discussion, let's assume that the video refresh rate is arbitrarily high, so that as soon as a frame becomes ready, we get to see it.) When a 3d card completes the rendering of a frame and swaps the front and back draw buffers, you are seeing the state of the world as it was at the time the game engine _began_ to draw the frame. If the current frame render time is x milliseconds, that's x milliseconds latency between the game state and your eyeballs.
With a spatial partitioning like SLI, both chipsets work in parallel to render a particular frame, and each frame is completed before rendering of the next frame begins; the game state to eyeball latency is simply 1/(frame rate).
With the ATI approach however, each of the two Rage chips plugs away at its frame independent of the other (which is working on a frame either one ahead or one behind.) Frame _render_ time is therefore twice the frame _display_ time, and the latency is twice as high as SLI for a given overall frame rate: 2/(frame rate). For a 60Hz frame rate, SLI gives 16.6 ms game state to eyeball latency, while the ATI approach gives 33.3 ms.
I am not a cognitive psychologist, so I don't know if an extra 16.6ms or so is going to make a noticeable difference to most people, but I wouldn't be suprised if experienced first-person-shooter players noticed a difference. Certainly for modem play the extra latency is probably smaller than the variation in ping time to the server, so I wouldn't expect it to make much difference, but on a LAN it might be noticeable. I have turned off sync to vertical refresh and forgone triple buffering in LAN Q3Test games because the variation in latency between frames was driving me batty, so I think this could actually be an issue. Of course, the higher the frame rate, the smaller the extra latency, and the less this will matter.
There is also the other matter that for this to work, there has to be at all times a large amount of rendering waiting to go so that each chipset stays busy. The drivers will presumably have to do a *lot* of buffering and then spoon feed each chip as its command FIFO is exhausted. I really wonder whether this will fit in well with what currently written applications are expecting from 3d acceleration hardware; if an application wants to have any synchronous interaction at all with the hardware, such as reading back values from a stencil buffer each frame after drawing is complete, it will totally screw this kind of pipelining. Somehow I'm just not convinced.
-Snorbert, somewhere in the antipodes
the Maxx technology on the card is a good idea but its not something I'm going to think about buying anytime soon just like the GeForce. The chips on this card basically suck, they run hot and don't have anything new but what they do have is a new chipset that has a future. The chipset was designed so when they come out with the next generation of their big and bad chips they can stick them in a Maxx chipset and double the framerate and maybe even performance. The GeForce is the same in my opinion, the idea is sound and will be cooler in a few months but right now it's a waste of money. I think chipmakers are all going to move to a GeForce styling where the video processor is a GPU (or whatever) that takes some of the strain off the system's CPU. My P3 500 with a Viper v770 plays any game I put on the machine damn well but in two or three years it most likely won't. If I want the newest games (assuming that video card technology stayed the same as it is now where the graphics card just handles the actual rendering and output) I would have to update my entire system which would mean I'd build a new one. Then along comes a technology like the GeForce where you can just get a new graphics board that deflates the load on the CPU by handling more parts of the game. I no longer have to completely update my system, I can keep it competitive by replacing the faster and more powerful graphics board. Just look at the games that ran great on a 90mhz Pentium processor a few years ago. If they had a way for more and more of the game to be run on the graphics board then your 90mhz pentium might be able to play Quake 2 pretty decently since it would only have to run the game code and not the graphics setup. Maybe in another year or two we'll see an entire gaming subsystem you could plug into your AGP slot, like having a Dreamcast in your computer. The subsystem could run every aspect of the games and use the rest of the parts of the PC as imput. The CPU wouldn't even have to be involved because DMA devices would mean they could communicate directly with the subsystem without bothering the tired CPU.
I'm a loner Dottie, a Rebel.
Tom's Hardware has a full preview, although he's not allowed to print the performance results. This isn't SLI or PGC -- ATI is actually having the CPU's draw full alternate frames, so the image quality will be high yet the speed will be doubled. It's even buffered so if CPU0 is taking a long time drawing frame 0, CPU1 can keep drawing 1,2,3,... until CPU0 is done. What I like best about this is that the MAXX architecture will allow them to drop their latest chips in as they are developed, so even if their CPU architectures remain a little behind, they'll be able to keep competitive. I like ATI cards because of all of the MPEG and TV toys they build into them; the only other company that even comes close to offering those kind of toys is Matrox, but they're just too damn expensive for the full-featured cards.
A XFCom X server for Linux is available for the single Rage 128 at Suse. I believe it will probably mainline in the next version of the XFree86 server, it seems quite stable. This should mean that most of the information for the dual processor version is already known, so hopefully a driver will follow quickly.
If you mean DRI and GL support, that's coming in version XFree86 v4.0, which hasn't announced a release date. The next snapshot, 3.9.17 should be available mid-month according to the XFree86 page.
Personally, I don't find a problem with the performance of my ATI Fury 32. I bought it because it had MPEG motion compensation and TV out and 32 megs of video ram, I've kept it because its fast enough for me, though I sort of hedged my bets by adding a voodoo 2 daughter card (passthrough) for those games when I really want it. The combo of the fury for 2d and some 3d stuff (especially some of the better GL software) and voodoo when I need it for certain games makes a pretty formidable video subsystem. Now, if I could only get X to work with the Fury 32 I wouldn't be wasting a PIII/500 with 256 Megs of Ram on wind*ws ;-)
I reject your reality
SharkyExtreme has a more lengthy writeup, including some initial performance comparisons from a prerelease version (chips clocked to 125Mhz instead of 143Mhz, beta drivers).
Does anyone here remember when www.ati.com was run by some company calling itself "Artificial Turd Industries"? The home page featured a very large, very detailed image of rubber doggie doo.
/. and I'm sure the fall of such a rebel would have been noticed...? Does anyone know if the guy finally sold out or if some how the courts decided that ATI should get the site (even though there are many other companies that have the same initials trademarked!)
= -=-=-=-=-=-=-=-=-=-
The thing that was neat was that this page stayed that way for as long as I could remember. The owner took great delight in posting letters from lawyers demanding he turn over the domain name. Companies like ATI Technologies (the graphics card maker most people are trying to find when they type in www.ati.com), American Tractor Incorporated, Arand Typeset and Ink...and about a dozen others.
ATI ended up getting www.atitech.com which they still own. But now I just found that they also have acquired www.ati.com!
How did this happen? I don't remember reading about it on
- JoeCurious
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
-- I wonder which will go down in history as the bigger failure: the War on Drugs or the War on Filesharing