Lots More Compute, a Leetle More Texturing

NVIDIA's GT200 GPU has a significant increase in computational power thanks to its 240 streaming processors, up from 128 in the previous G80 design. As a result, NVIDIA's GT200 GPU showcases a tremendous increase in transistor count over its previous generation architecture (1.4 billion up from 686 million in G80).

The increase in compute power of GT200 is not mirrored however in the increase in texture processing power. On the previous page we outlined how the Texture/Processing Clusters went from two Shader Multiprocessors to three, and how there are now a total of ten TPCs in the chip up from 8 in the GeForce 8800 GTX.

In the original G80 core, used in the GeForce 8800 GTX NVIDIA's texture block looked like this:

In each block you had 4 texture address units and 8 texture filtering units.

With the move to G92, used in the GeForce 8800 GT, 8800 GTS 512 and 9800 GTX, NVIDIA doubled the number of texture address units and achieved a 1:1 ratio of address/filtering units:

With GT200 in the GeForce GTX 280/260, NVIDIA kept the address-to-filtering ratio at 1:1 but increased the ratio of SPs to texture processors:

In the previous designs you'd have 8 address and 8 filtering units per TPC (or 16 streaming processors), in the GT200 you have the same 8 address and 8 filtering units but for a larger TPC with 24 SPs.

Here's how the specs stand up across the generations:

 NVIDIA Architecture Comparison G80 G92 GT200
Streaming Processors per TPC 16 16 24
Texture Address Units per TPC 4 8 8
Texture Filtering Units per TPC 8 8 8
Total SPs 128 128 240
Total Texture Address Units 32 64 80
Total Texture Filtering Units 64 64 80

 

For a 87.5% increase in compute, there's a mere 25% increase in texture processing power. This ratio echoes what NVIDIA has been preaching for years: that games are running more complex shaders and are not as bound by texture processing as they were in years prior. If this wasn't true then we'd see a closer to 25% increase in performance of GT200 over G80 at the same clock rather than something much greater.

It also means that GT200's performance advantage over G80 or G92 based architectures (e.g. GeForce 9800 GTX) will be determined much by how computationally bound the games we're testing are.

The ratio of increase compute/texture power in the GT200 has been evident in NVIDIA architectures for years now, dating back to the ill-fated GeForce FX. NVIDIA sacrificed memory bandwidth on the GeForce FX, equipping it with a narrow 128-bit memory bus (compared to ATI's 256-bit interface on the Radeon 9700 Pro) and instead focused on building a much more powerful compute engine. Unfortunately, the bet was the wrong one to make at the time and the GeForce FX was hardly competitive (for more reasons than just a lack of memory bandwidth), but today we're dealing in a very different world. Complex shader programs run on each pixel on the screen and there's a definite need for more compute power in today's GPUs.

An Increase in Rasterization Throughput

In addition to the 25% increase in texture processing capabilities of the GT200, NVIDIA added two more ROP partitions to the GPU. While the GeForce 8800 GTX had six ROP partitions, each capable of outputting a maximum of 4 pixels per clock, the GT200 adds two more partitions.

With eight ROP partitions the GT200 can now output a maximum of 32 pixels per clock, up from 24 pixels per clock in the GeForce 8800 GTX and 9800 GTX.

The pixel blend rate on G80/G92 was half-speed, meaning that while you could output 24 pixels per clock, you could only blend 12 pixels per clock. Thanks to the 65nm shrink and redesign, GT200 can now output and blend pixels at full speed - that's 32 pixels per clock for each.

The end result is a non-linear performance improvement in everything from anti-aliasing and fire effects to shadows on GT200. It's an evolutionary change, but that really does sum up many of the enhancements of GT200 over G80/G92.

Building NVIDIA's GT200 Derek Gets Technical: 15th Century Loom Technology Makes a Comeback
POST A COMMENT

108 Comments

View All Comments

  • skiboysteve - Tuesday, June 17, 2008 - link

    FANTASTIC write up on fine-grained TMT. I was unaware about this threading technique and was always thinking of this in class or whenever someone would talk about hyperthreading. this technique was literaly in my head for well over a year and I didn't know what it was called or that it even had a name. I always thought there had to be a more elegant way than hyperthreading to do multithreading down at the chip level without doing the OS style time slicing.

    i was sitting there wondering how the hell the schedule and run these SPs and then bam whole page about it

    really appreciate the effort that goes into researching the core of these chips. i know not everyone likes it but for guys that are educated and work in the field its really interesting
    Reply
  • DerekWilson - Tuesday, June 17, 2008 - link

    remember though that this type of fine-grained TMT only has payoffs in systems running millions of threads concurrently.

    on an OS you'll see hundreds or even thousands of threads on heavily used systems, but there still wouldn't be enough concurrent action to justify this type of architecture for general purpose computing.

    of course, as developers push towards an effort to thread their code as much as possible, who knows what architectures might be worth exploring on the desktop ...
    Reply
  • coder0000 - Tuesday, June 17, 2008 - link

    Very well written! A couple of points:

    1) Last week at WWDC Apple announced OpenCL as an alternative to CUDA. It's a C99 based HLL for creating compute kernels that can be deployed to GPU's and CPU's. Today Khronos officially announced a working group for this, and NV is a part of the committee. As such, your wish for an industry standardized compute language similar to CUDA that runs on all platforms and vendors HW may not be so far off.

    2) I believe your interpretation of how multiple threads simultaneously execute in an SM is incorrect. Per thread context switching is not free, and you would never be able to execute a different thread every cycle in the manner described. There is far too much context that needs to be swapped out, and there would be significant power implications for doing that, in addition to the latency. Instead, I believe what NV is claiming is that any given SP executes a single thread. All threads in the SM can all be a single warp, but you can also have multiple threads (one per SP) all executing simultaneously in an SM.
    Reply
  • DerekWilson - Tuesday, June 17, 2008 - link

    1) I haven't had a good chance to look at OpenCL, but I certainly hope that if it's everything everyone is saying it is in the comments here that it takes off in a bigger way than CUDA :-)

    2) it does not context switch per thread -- warps define a context, and you have 32 threads grouped together. these threads all share the same instruction stream, which is why if threads in a warp take different directions on a branch all 32 threds must follow both paths.

    NVIDIA has flat out stated that every schedule clock a new warp is scheduled and that it takes 4 clock cycles to process one warp on an SM. For both of these to be true, we conclude that the scheduler alternates scheduling SPs and SFUs on altenating clocks which means the SPs would be scheduled every 4 clocks relative to itself.

    On 8 SPs per SM, you some how need to execute 32 threads in 4 clock cycles. This makes sense if you execute 4 threads per SP in some way. The details at this point are fuzzy though.

    regardless, if an SP executes 4 different threads from the same warp, there is no need to context switch to execute any of these threads -- again, threads in the same warp share context.
    Reply
  • skiboysteve - Tuesday, June 17, 2008 - link

    could be a large explanation of the 2x register file size. and remember that the SP doesn't have to worry about the context switch, the SM handles having the data in the right place Reply
  • anandtech02148 - Monday, June 16, 2008 - link

    From this conclusion, Amd seems to be the shrewd player, let nvidia and intel duke it out in the high voltage, heat, meaningless speed gpu while Amd can pull something like its first dualcore or athlon64 for the win.
    this new beast from Nvidia will have how many developers making games for it right away? i'm guestimating maybe 2yrs-4yrs down the road we'll see a decent title that take full advantage of this hardware.
    by then Amd will have something of a midrange that can more than handle the games.
    2 things nvidia could work on that it already has, the ps3 market, and small graphic devices to improve profits. shrink the ps3 gpu further so Sony can shrink it's machinel and sell more.

    Reply
  • PrinceGaz - Monday, June 16, 2008 - link

    The GT200 core may be a technical masterpeice in terms of actually making something that big which is fully functional on GTX280 cards, but it seems to me the penalty of fabbing it at 65nm negates much of the benefits of such a wide GPU.

    They've had to drop the clock speeds throughout presumably because of the ridiculous amount of heat such a large core generates, which means the ~60% performance advantage in current games over the G80 core at similar clock-speeds is somewhat reduced.

    Given that ATI are not producing their 55nm cores in AMD's fabs but instead are getting them churned out reliably elsewhere, nVidia have made a mistake this time around in having their high-end product rely on previous-generation fabrication as it makes it run too hot to allow the clock-speeds needed for it to be the product it should be. There is always a risk in transitioning to a smaller fab technology, and nVidia suffered badly in the past by doing so too early, but with a chip the size of the GT200, they really should have gone to 55nm even if it meant a delay of a month or three, whilst the smaller cut-down derivatives were rolled out first.
    Reply
  • ekpyr - Monday, June 16, 2008 - link

    Great article, but what about the microstuttering issues present in Nvidia's 9800GX2 cards (both SLI and Quad-SLI)? There is very little discussion on this, but I've seen some benchmarks where the FPS floor is 4fps with the 9800GX2s. Can you add a subjective review of whether or not the actual gameplay is smoother with the GTX280s across these games? Aggregate numbers may say one thing, but I've returned a 9800 GX2 Quad-SLI setup because it was unable to handle the incredible amount of texture loading that was done in Age of Conan (2560x1600 4xAA 'High' settings = 4fps). The 8800 GTX Tri-SLI configuration I am currently using is more resilient to microstuttering with its increased bus and memory capacities, but I'm very curious about the GTX280s and their increased memory and bus on texture-heavy games like Age of Conan. Reply
  • DerekWilson - Monday, June 16, 2008 - link

    the only game that came close to having this issue with quad sli for us was oblivion.

    in that game at high res lag and stutter are unbearable and the game is unplayable.

    we didn't notice any stuttering issues with a single GX2.

    i'm working on some analysis tools to show details like this better in future articles.
    Reply
  • TheJian - Monday, June 16, 2008 - link

    I find it humorous that nobody discusses the fact that the shrink has already taped out and will likely be out in two months or just after. This humongous chip was only released so that when AMD releases in the next few weeks they will be behind still in single GPU cards. This is basically what Intel does to AMD every time AMD has a better chip. For all intents and purposes this is a PAPER release of what will come in 2-2.5 months (In Intel's case they just show you what will be out 6 months from now, and a large portion of people don't buy an AMD because Intel might be ahead by xmas...LOL - works like a charm every time AMD is ahead). THE DIE SHRUNK CHIP! Most likely with faster speeds. I suspect they'll come with "ULTRA" version first (and stick it on top of the price heap, so as to not kill all FAT cards in the channel already) and then filter down as these big suckers leave the channel. That's if they even plan to sell more than a few of these to begin withat 65nm. It's only out there so AMD won't look any good in two weeks.

    MIND SHARE is everything, which is why Intel's KING of the paper launch when behind strategy. They've even went to doing it for all chips no matter what now. Nehalem scores 6 months before availability. AMD's marketers have no clue an should be fired. You have to play the same DIRTY game as your enemy or you've already lost. If AMD had half a brain in their head they'd paper launch an ultra or 2x4870 version for the same reason...LOL. Then claim "our 4870x2 makes nvidia look like crap for $600"...ROFL. Who cares when it's available, just say it. Having said that, Nvidia will wipe the floor with them in 2 months anyway on a 2xGTX280 that's die shrunk. Which is all they are doing today...BUYING TIME!
    Reply

Log in

Don't have an account? Sign up now