ATI Radeon X800 Pro and XT Platinum Edition: R420 Arrives
by Derek Wilson on May 4, 2004 10:28 AM EST- Posted in
- GPUs
Pixel Shader Performance Tests
ShaderMark v2.0 is a program designed to stress test the shader performance of modern DX9 graphics hardware with Shader Model 2.0 programs written in HLSL running on a couple shapes in a scene.
We haven't used ShaderMark in the past because we don't advocate the idea of trying to predict the performance of real world game code using a synthetic set of tests designed to push the hardware. Honestly, as we've said before, the only way to determine performance of a certain program on specific hardware is to run that program on that hardware. As both software and hardware get more complex, results of any given test become less and less generalize able, and games, graphics hardware, and modern computer systems are some of the most complex entities on earth.
So why are we using ShaderMark you may ask. There are a couple reasons. First this is only a kind of ball park test. ATI and NVIDIA both have architectures that should be able to push a lot of shader operations through. It is a fact that NV3x had a bit of a handicap when it came to shader performance. A cursory glance at ShaderMark should tell us enough to know if that handicap carries over to the current generation of cards, and whether or not R420 and NV40 are on the same playing field. We don't want to make a direct comparison, we just want to get a feel for the situation. With that in mind, here are the benchmarks.
| Radeon X800 XT PE | Radeon X800 Pro | GeForce 6800 Ultra | GeForce 6800 GT | GeForce FX 5950 U | |
|---|---|---|---|---|---|
| 2 | 310 |
217 |
355 |
314 |
65 |
| 3 | 244 |
170 |
213 |
188 |
43 |
| 4 | 238 |
165 |
|||
| 5 | 211 |
146 |
162 |
143 |
34 |
| 6 | 244 |
169 |
211 |
187 |
43 |
| 7 | 277 |
160 |
205 |
182 |
36 |
| 8 | 176 |
121 |
|||
| 9 | 157 |
107 |
124 |
110 |
20 |
| 10 | 352 |
249 |
448 |
410 |
72 |
| 11 | 291 |
206 |
276 |
248 |
54 |
| 12 | 220 |
153 |
188 |
167 |
34 |
| 13 | 134 |
89 |
133 |
118 |
20 |
| 14 | 140 |
106 |
141 |
129 |
29 |
| 15 | 195 |
134 |
145 |
128 |
29 |
| 16 | 163 |
113 |
149 |
133 |
27 |
| 17 | 18 |
13 |
15 |
13 |
3 |
| 18 | 159 |
111 |
99 |
89 |
17 |
| 19 | 49 |
34 |
|||
| 20 | 78 |
56 |
|||
| 21 | 85 |
61 |
|||
| 22 | 47 |
33 |
|||
| 23 | 49 |
43 |
49 |
46 |
These benchmarks are run with fp32 on NVIDIA hardware and fp24 on ATI hardware. It isn't really an apples to apples comparison, but with some of the shaders used in shadermark, partial precision floating point causes error accumulation (since this is a benchmark designed to stress shader performance, this is not surprising).
ShaderMark v2.0 clearly shows huge increase in pixel shader performance from NV38 to either flavor of NV40. Even though the results can't really be compared apples to apples (because of the difference in precision), NVIDIA manages to keep up with the ATI hardware fairly well. In fact, under the diffuse lighting and environment mapping, shadowed bump mapping and water color shaders don't show ATI wiping the floor with NVIDIA.
In looking at data collected on the 60.72 version of the NVIDIA driver, no frame rates changed and a visual inspection of the images output by each driver yielded no red flags.
We would like to stress again that these numbers are not apples to apples numbers, but the relative performance of each GPU indicates that the ATI and NVIDIA architectures are very close to comparable from a pixel shader standpoint (with each architecture having different favored types of shader or operation).
In addition to getting a small idea of performance, we can also look deep into the hearts of NV40 and see what happens when we enable partial precision rendering mode in terms of performance gains. As we have stated before, there were a few image quality issues with the types of shaders ShaderMark runs, but this bit of analysis will stick only to how much work is getting done in the same amount of time without regard to the relative quality of the work.
| GeForce 6800 U PP | GeForce 6800 GT PP | GeForce 6800 U | GeForce 6800 GT | |
|---|---|---|---|---|
| 2 | 413 |
369 |
355 |
314 |
| 3 | 320 |
283 |
213 |
188 |
| 5 | 250 |
221 |
162 |
143 |
| 6 | 300 |
268 |
211 |
187 |
| 7 | 285 |
255 |
205 |
182 |
| 9 | 159 |
142 |
124 |
110 |
| 10 | 432 |
389 |
448 |
410 |
| 11 | 288 |
259 |
276 |
248 |
| 12 | 258 |
225 |
188 |
167 |
| 13 | 175 |
150 |
133 |
118 |
| 14 | 167 |
150 |
141 |
129 |
| 15 | 195 |
173 |
145 |
128 |
| 16 | 180 |
161 |
149 |
133 |
| 17 | 21 |
19 |
15 |
13 |
| 18 | 155 |
139 |
99 |
89 |
| 23 | 49 |
46 |
49 |
46 |
The most obvious thing to notice is that, overall, partial precision mode rendering increases shader rendering speed. Shader 2 through 8 are lighting shaders (with 2 being a simple diffuse lighting shader). These lighting shaders (especially the point and spot light shaders) will make heavy use of vector normalization. As we are running in partial precision mode, this should translate to a partial precision normalize, which is a "free" operation on NV40. Almost any time a partial precision normalize is needed, NV40 will be able to schedule the instruction immediately. This is not the case when dealing with full precision normalization, so the many 50% performance gains coming out of those lighting shaders is probably due to the partial precision normalization hardware built into each shader unit in NV40. The smaller performance gains (which, interestingly, occur on the shaders that have image quality issues) are most likely the result of decreased bandwidth requirements, and decreased register pressure: a single internal fp32 register can handle two fp16 values making scheduling and managing resources much less of a task for the hardware.
As we work on our image quality analysis of NV40 and R420, we will be paying heavy attention to shader performance in both full and partial precision modes (as we want to look at what gamers will actually be seeing in the real world). We will likely bring shadermark back for these tests as well. This is a new benchmark for us, so please bear with us as we get used to its ins and outs.
Facebook
Twitter
RSS
95 Comments
View All Comments
Pumpkinierre - Wednesday, May 5, 2004 - link
Sorry, scrub that last one. I couldnt help it. I will reform.Pumpkinierre - Wednesday, May 5, 2004 - link
So, which is better: a64 at 2Gig or P4 at 3.2?jibbo - Wednesday, May 5, 2004 - link
"Zobar is right; contra Jibbo, the increased flexibility of PS3 means that for many 2.0 shader programs a PS3 version can achieve equivalent results with a lesser performance hit."I think you're both still missing my point. There is nothing that says PS3.0 is faster than PS2.0. You are both correct that it has to potential to be faster, though you both assume that a first generation PS3.0 architecture will perform at the same level as a refined PS2.0 architechture.
PS3.0 is one of the big reasons that nVidia's die size and transistor count are bigger than ATI's. The additional power drain (and consequently heat dissipation) of those 40M transistors also helps to limit the clock speeds of the 6800. When you're talking about ALU ops per second (which dominate math-intensive shaders), these clock speeds become very important. A lot of the 6800's speed for PS3.0 will have to be found in the driver optimizations that will compile these shaders for PS3.0. Left to itself, ATI's raw shader performance still slaughters nVidia's.
They both made trade-offs, and it seems that ATI is banking that PS3.0 won't be a dealbreaker in 2004. Only time will tell....
Phiro - Wednesday, May 5, 2004 - link
K, I found the $400M that the CEO claimed. He also claimed $400M for the NV3x core as well. It seemed more as a boast than anything, not particularly scientific or exact.In any case, ATI supposedly spent $165-180M last year (2003) on R&D, with an estimated increase of 100% for this year. How long has the 4xx core been in development?
Regardless, ultimately we the consumers are the winners. Whether or not the R&D spent pans out will play out over the next couple years, as supposedly the nv4x core has a 24 month lifespan.
413xram - Wednesday, May 5, 2004 - link
If you watch nvidia's launch video on their site they mention the r&d costs for their new card.RyanVM - Wednesday, May 5, 2004 - link
What ever happened to using ePSXe as a video card benchmark?Phiro - Wednesday, May 5, 2004 - link
Well, Nvidia may have spent $400M on this (I've never seen that number before but we'll go with it I guess) but they paid themselves for the most part.ATI's cost can't be too trivialized - didn't they drop a product design or two in favor of getting this out the door instead? And any alteration in the architecture of something doesn't really qualify as a hardware "refresh" in my book - a hardware refresh for an OEM consists of maybe one speed notch increase in the RAM, new bios, larger default HD, stuff like that. MLK is what Dell used to call it - Mid Life Kick.
retrospooty - Wednesday, May 5, 2004 - link
"Precisely. By the time 512mb is useful, the card will be too slow for it to matter, and you'd need a new card any way."True...
Both cards perform great, both have wins and losses depending on the game. The deciding factor will be price and power requirements.
Since prices will adjust downward, at a fairly equal rate, that leaves power. With Power requirements being so incredibly high with the NV40, that leans me toward ATI.
413xram also has a good point above. For Nvidia, this is a 400 million dollar new chip design. For ATI, this was a refresh of an old design to add 16 pipes, and a few other features. After the losses NV took with the heavily flawed NV30 and 35 , they need a financial boom, and this isnt it.
mattsaccount - Wednesday, May 5, 2004 - link
There are no games available today that use 256mb of video RAM, let alone 512mb. Even upper-high-end cards routinely come with 128mb (e.g. Geforce FX 5900, Radeon 9600XT). It would not make financial sense for a game developer to release a game that only a small fraction of the community could run acceptably.>> I have learned from the past that future possibilties of technology in hardware does nothing for me today.
Precisely. By the time 512mb is useful, the card will be too slow for it to matter, and you'd need a new card any way.
413xram - Wednesday, May 5, 2004 - link
#64 Can you explain "gimmick"?