The Pixel Shader Engine

On par with what we have seen from NVIDIA, ATI's top of the line card is offering a GPU with a 16x1 pixel pipeline architecture. This means that it is able to render up to 16 single textured pixels in parallel per clock. As previously alluded to, R420 divides its pixel pipes into groups of four called quads. The new line of ATI GPUs will offer anywhere from one to four quad pipelines. The R3xx architecture offers an 8x1 pixel pipeline layout (grouped into two quad pipelines), delivering half of R420's pixel processing power per clock. For both R420 and R3xx, certain resources are shared between individual pixel pipelines in each quad. It makes a lot of sense to share local memory among quad members, as pixels near eachother on the screen should have (especially texture) data with a high locality of reference. At this level of abstraction, things are essentially the same as NV40's architecture.

Of course, it isn't enough to just look how many pixel pipelines are available: we must also discover how much work each pipeline is able to get done. As we saw in our final analysis of what went wrong with NV3x, the internals of a shader unit can have a very large impact on the ability of the GPU to schedule and execute shader code quickly and efficiently.

At our first introduction, the inside of R420's pixel pipeline was presented as a collection of 2 vector units, 2 scalar units, and one texture unit that can all work in parallel. We've seen the two math and one texture layout of NV40's pixel pipeline, but does this mean that R420 will be able to completely blow NV40 out of the water? In short, no: it's all about what kind of work these different units can do.

Lifting up the hood, we see that ATI has taken a different approach to presenting their architecture than NVIDIA. ATI's presentation of 2 vector units (which are 3 wide at 72bits), 2 scalar units (24bits), and a texture unit may be more reflective of their implementation than what NVIDIA has shown (but we really can't know this without many more low level details). NVIDIA's hardware isn't quite as straight forward as it may end up looking to software. The fact is that we could look at the shader units in NV40's pixel pipeline in the same way as ATI's hardware (with the exception of the fact that the texture unit shares some functionality with one of the math units). We could also look at NV40 architecture as being 4 2-wide vector units or 2 4-wide vector units (though this is still an over simplification as there are special cases NVIDIA's compiler can exploit that allow more work to be done in parallel). If ATI had decided to present it's architecture in the same way as NVIDIA, we would have seen 2 shader math units and one completely independent texture unit.

In order to gain better understanding, here is a diagram of the parallelism and functionality of the shader units within the pixel pipelines of R420 and NV40:


ATI has essentially three large blocks that can push up to 5 operations per clock cycle


NV40 can be seen two blocks of a more amorphous configuration (but there are special cases that allow some of these parts to work at the same time within each block.

Interestingly enough, there haven't been any changes to the block diagram of a pixel pipeline at this level of detail from R3xx to R420.

The big difference in the pixel pipe architectures that gives the R420 GPU a possible upper hand in performance over NV40 is that texture operations can be done entirely in parallel with the other math units. When NV40 needs to execute a texture operation, it looses much of its math processing power (the texturing unit cannot operate totally independently of the first shader unit in the NV40 pixel pipeline). This is also a feature of R3xx that carried over to R420.

Understanding what this all means in terms of shader performance depends on the kind of code developers end up writing. We wanted to dedicate some time to hand mapping some shader code to both architecture's pixel pipelines in order to explain how each GPU handled different situations. Trial and error have led us to the conclusion that video card drivers have their work cut out for them when trying to optimize code; especially for NV40. There are multiple special cases that allow NVIDIA's architecture to schedule instructions during texturing operations on the shared math/texture unit, and some of the "OR" cases from our previous diagram of parallelism can be massaged into "and" cases when the right instructions are involved. This also indicates that performance gains due to compiler optimizations could be in NV40's future.

Generally, when running code with mixed math and texturing (with a little more math than texturing) ATI will lead in performance. This case is probably the most indicative of real code.

The real enhancements to the R420 pixel pipeline are deep within the engine. ATI hasn't disclosed to us the number of internal registers their architectures have, or how many pixels each GPU can maintain in flight at any given time, or even cache hit/miss latencies. We do know that, in addition to the extra registers (32 constant and 32 temp registers up from 12) and longer length shaders (somewhere between 512 and 1536 depending on what's being done) available to developers on R420, the number of internal registers has increased and the maximum number of pixels in flight has increased. These facts are really important in understanding performance. The fundamental layout of the pixel pipelines in R420 and NV40 are not that different, but the underlying hardware is where the power comes from. In this case, the number of internal pipeline stages in each pixel pipeline, and the ability of the hardware to hide the latency of a texture fetch are of the utmost importance.

The bottom line is that R420 has the potential to execute more PS 2.0 instructions per clock than NVIDIA in the pixel pipeline because of the way it handles texturing. Even though NVIDIA's scheduler can help to allow more math to be done in parallel with texturing, NV40's texture and math parallelism only approaches that of ATI. Combine that with the fact that R420 runs at a higher clock speed than NV40, and even more pixel shader work can get done in the same amount of time on R420 (which translates into the possibility for frames being rendered faster under the right conditions).

Of course, when working with fp32 data, NV40 is doing 25% more "work" per operation, and it's likely that the support for fp32 from the front of the shader pipeline to the back contributes greatly to the gap in the transistor count (as well as performance numbers). When fp16 is enabled in NV40, internal register pressure is decreased, and less work is being done than in fp32 mode. This results in improved performance for NV40, but questions abound as to real world image quality from NVIDIA's compiler and precision optimized shaders (we are currently exploring this issue and will be following up with a full image quality analysis of now current generation hardware).

As an extension of the fp32 vs. fp24 vs. fp16 debate, NV40's support of Shader Model 3.0 puts it at a slight performance disadvantage. By supporting fp32 all the way through the shader pipeline, flow control, fp16 to the framebuffer and all the other bells and whistles that have come along for the ride, NV40 adds complexity to the hardware, and size to the die. The downside for R420 is that it now lags behind on the feature set front. As we pointed out earlier, the only really new features of the R420 pixel shaders are: higher instruction count shader programs, 32 temporary registers, and a polygon facing register (which can help enable two sided lighting).

To round out the enhancements to the R420's pixel pipeline, ATI's F-Buffer has been tweaked. The F-Buffer is what ATI calls the memory that stores pixels that have come out of the pixel shader but still require another pass (or more) thorough the pixel shader pipeline in order to finish being processed. Since the F-Buffer can require anywhere from no memory to enough memory to handle every pixel coming down the pipeline, ATI have built "improved" memory management hardware into the GPU rather than relegating this task to the driver.

The R420 Vertex Pipeline Depth and Stencil with Hyper Z HD
Comments Locked

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"?

Log in

Don't have an account? Sign up now