Slashdot Mirror
Add Another Core for Faster Graphics
Dzonatas writes "Need a reason for extra cores inside your box? How about faster graphics. Unlike traditional faster GPUs, raytraced graphics scale with extra cores. Brett Thomas writes in his article Parallel Worlds on Bit-Tech, 'But rather than working on that advancement, most of the commercial graphics industry has been intent on pushing raster-based graphics as far as they could go. Research has been slow in raytracing, whereas raster graphic research has continued to be milked for every approximate drop it closely resembles being worth. Of course, it is to be expected that current technology be pushed, and it was a bit of a pipe dream to think that the whole industry should redesign itself over raytracing.' A report by Intel about Ray Tracing shows that a single P4 3.2Ghz is capable of 100 million raysegs, which gives a comfortable 30fps. Intel further states 450 million raysegs is when it gets 'interesting.' Also, quad cores are dated to be available around the turn of the year. Would octacores bring us dual screen or separate right/left real-time raytraced 3D?"
Need a reason for extra cores inside your box? No :)
-- "Genius is 1% inspiration and 99% perspiration" - TAE --
Lemme see, at this rate I'll need: 9 cores for the raytracer, 7 cores for the physics simulation, 5 for the AI, 3 for the OS, and of course
;)
One core to rule them all
One core to find them
One core to bring them all
And in the darkness bind them
A polar bear is a cartesian bear after a coordinate transform.
There are already ray traced games. :O
http://graphics.cs.uni-sb.de/~morfiel/oasen/
F.A.N. released a real-time raytraced demo at breakpoint back in 2003. It does no more than 10 fps on my lowly 1GHz P3, but I'm sure it runs quite smooth on a nice modern CPU (though I don't think it's multithreaded).
Error: password can't contain reverse spelling of ancient Chinese emperor
Each core is already capable of doing 100 million raysegs and you talk about quad cores. So I think you mean
450 million raysegs not 450 raysegs.
Slashdot: Tabloid for the nerds. Stuff that doesn't matter.
Take a look at the proceedings from any graphics conference in the last three or four years, and you will see several papers which involve ray-tracing on a GPU. Actually, not so many recently, because it's been done to death. The most impressive one I saw was at Eurographics in 2004 running non-linear ray tracing. As the rays advanced, their direction was adjusted based on the gravity of objects in the scene. The demo (rendered in realtime) showed a black hole moving in front of a stellar scene and all of the light effects this caused.
I am TheRaven on Soylent News
If i remember it correctly from my days of playing with POVRay (free raytracing app), the time it took to raytrace an image depended on things like the presence (or not) of semi-transparent, semi-reflective surfaces and on the number of light sources.
If this is still the case, then going from the current rendering techniques in games to raytracing would result in images with more realistic reflections and lighting but, due to performance tradeoffs, few reflective surfaces and light sources.
Besides, at the moment what games need the most is beter AIs and procedurally generated content, not yet another layer of eyecandy that requires gamers to upgrade their hardware (again).
FTA
"Oh, blast. Rabbit, I seem to have forgotten my pocketwatch. May I borrow yours?"
Rabbit: I'm late, I'm late, I'm late...
---
anyway, if these technology becomes a reality in the 3-5 years and if I read the article right, the whole graphics architecture would change, there would only be a need for a super graphics processor and less need for too much memory and those graphics pipeline/shader thingies...
The reason that they might want it in a CPU is that, why have a separate add on GPU to handle the job while the CPU could do it alone by that time. You would then only need a "basic" video card that would just do the display.
Hmmm... could this be one of the reasons why ATI and AMD merged?
Just found that game using raytracing - Quake 3: Raytraced./
http://graphics.cs.uni-sb.de/~sidapohl/egoshooter
Rumors are there's a q4 version on the way.
I wonder how much this research relates to Intel's renewed desire to become a graphics player.
If they're having trouble, for staffing or other reasons, producing good GPU designs, then it would be pretty clever of them to revolutionize the industry AND capitalize on their CPU strengths in a single move. More power to them, I say. (More power = about 120 watts, I'm guessing.)
No, ray tracing is all about searching databases for ray-object intersections. That's what GPUs can't do at all.
No sig today...
It's not just the sheer number of FP calculations that can be the problem. Once you get away from the first (or perhaps even second) level of rays, you end up losing coherence between neighbouring rays which causes memory page/cache thrashing. This is not a nice thing on a GPU.
It's extremely unlikely that anything will go anywhere with raytracing in the near future. Raytracing takes a tremendous amount of power - apps that demonstrate it in realtime usually run quite choppy, and they're very minimalistic to boot; ugly textures, very simple geometry, very confined areas...
The main benefits of raytracing in games would be:
1) Shadows; they'd be Doom 3-like. Several games have full stencil shadows and that's just how raytraced ones would look: sharp and straight. The difference? Raytraced ones would take a ton more power and time to compute.
2) True reflection and refraction. We can "fake" this well enough - for example, see the Source engine's water, incorporating realtime fresnel reflections and refractions. Though Source's water's "fake" refraction/reflection aren't pixel-perfect, and are only distorted by a bump-map, it certainly looks great.
Honestly, considering the small gain in visual quality (although a major gain in accuracy) - it's like going after a fly with a bazooka. Sure, once we get to the point where there's enough processing power to deal with this well enough in realtime, it will happen - but don't expect it soon, and don't expect that huge a difference. Nicer reflections and refractions (which already look good today) and pixel-perfect shadows (looking just the same as stencil shadows in some newer games).
Ray tracing also suffers terribly from "jaggies". Edges look bad because rays can just miss an object and cause really bad stepping on the edges of objects. To eliminate jaggies and do anti-aliasing, you need to do sub-pixel rendering with jitter (slight randomness) to produce an average value for the pixel. So you might have to trace 4 or more rays in a pixel for acceptable anti-aliasing. Effects like focal length, fog, bump mapping etc. cause things to get even more complex. Most pictures rendered with high quality on Blender, POVRay etc. would take minutes if not hours even on a fast / dual core processor.
The only way you'd get 30fps is if cut your ray trace depth to 1 or 2, used a couple of lights, cut the screen res down and forgot about fixing jaggies. It would look terrible. Oh and find time for all the other things that apps and games must do.
A ray-tracing problem can be solved simultaneously using a moment method that incorporates physical optics. I wrote my Master's thesis a long time ago that did precisely that for 2-dimensional situations. Of course, this required solving massive linear systems that, at the time I wrote it, took hours on a 433MHz Alpha to do a single frame, and it was written in FORTRAN77, but hey, we've come a long way since then :)
I've seen the topic of realtime ray-tracing and hardware accelerated ray-tracing come up countless times over the past 15 years. In the 80's and 90's, a realtime ray-tracing acceleration chip was always around the corner. Some products did actually emerge, but never quite caught on. The reason for this is not because "commercial graphics industry has been intent on pushing raster-based graphics as far as they could go". Quite the contrary; it's much more elegant algorithmically (and hence 'easier') to implement a ray-tracer than a scanline based renderer. However, there's a fundamental limitation of ray-tracing that make it unappealing performance-wise. Cache coherence for ray-tracers suck.
All rendering algorithms boil down to a sorting problem, where all the geometry in the scene is sorted in the Z dimension per pixel or sample. Fundamentally, scanline algorithms and ray-tracing algorithms are the same. For primary rays, here's some simpliefied pseudocode:
foreach pixel in image
trace ray through pixel
shade frontmost geometry
The trace essentially sorts all the geometrty along its path.
A scanline algorithm looks like this:
foreach geometry object in the scene
foreach pixel geometry is in
if geometry is in front of whatever is in the pixel already
shade fragement of geometry in pixel
replace pixel with new shaded fragment
As you can see, the only distinction is the order of the two loops. For ray-tracing, traversing the pixels is in the outer loop, and the geometry in the inner loop. For scanline rendering, it's the opposite. This has huge consequences in terms of cache coherency. With scanline methods, since the same object is being shaded in the inner loop, and neighboring fragments of the same object are being shaded, cache coherency tends to be extermely high. The same shader program is used, and likelyhood of the texture being accessed from cache is very good. The same can't be said for ray-tracing. You can shoot two almost identical rays but touch wildly different parts of the scene. Cache coherency relative to scanline rendering is abysmal.
This one performance side-effect of ray-tracing is the only reason we haven't seen any serious ray-tracing for realtime applications. Even in offline rendering, scanline rendering dominates even though software ray-tracing has been available from the beginning of CG. For ray-tracing to become viable, we need more than just more CPU cores. We need buses fast enough to feed all the cores in situations where we have an extremely high ratio of cache misses. Unfortunately, the speed gap between memory speeds and compute power seems to be increasing in recent years.
Extra, extra! This just in! Report from CPU vendor discovers that you should spend more money on your CPU and less on your graphics card!
Shocking, I tells ya. Shocking.
Read my blog.
It's nice that people are working on ray traced games, but please note the following:
...
Oasen is based on "OpenRT" --- which is entirely proprietary, and is NOT open source. Their FAQ explains that clearly.
I'm sure that I'm not the only person annoyed at their use of "open" to mean "closed".
Time to look for an open-source raytracing engine designed for interative use
I am not an expert in OSes, I thought that the idle process just gave the CPU something to do while it waited for a working process (the idle just allowed the working to butt-in, whenever somethin came along).
Wouldn't creating a core just to do nothing be hardware bloat at its most obsurd?
Or am I showing my ignorance, just a bit too openly.
If this were really happening, what would you think?
1) Static Objects Only. The huge majority of computation time is traversing a spatial subdivision structure. It happens that K-d trees offer the best characteristic (typically, fewest primitive per leaf for a given memory limit). However, these are really heinous to dynamically update. You can cheaply re-create it with median partitioning, but your trees are crappy. You can do a much nicer SAH (surface area heuristic), but to do this per frame blows out your CPU budget.
2) Bandwidth. Even if you could update your subdivision structure very cheaply, that structure still needs to be propogated out to all the CPUs participating in the raytrace. For the 1.87 MTri model they list on page 6, their spatial structure was 127 MB. Say you have a bandwidth of 6 GB/s, it takes 20ms just to transfer the structure (and there are other problems here). So your ceiling is 50 Fps before you trace your first ray.
3) Slower than a GPU. Even though they give you some little graph showing that raytracing (a static model, with static partitioning) beats a GPU at a MTri in the frame, this is very deceiving. The GPU pipeline works such that zillions of sub-pixel triangles simply can't get into pixel shaders fast enough, and force the pixel shader to be run many times extra. Double the resolution, however and the GPU won't take a cycle longer... with raytracing, performance will halve. So they found a bottleneck in the GPU which is totally unrepresentative of a game in every single sense, and said LOOK! BETTER! (in theory).
4) Hey, Where's my Features? All the cool things about raytracing (nice shadows, refraction, implicit surfaces, reflection, subsurface scattering) all get tossed out the window to make it real-time! What's the point, then? Given all the pixel shader hacks invented to make a GPU frame look interesting, the quality that can be achieved in a real-time raytrace is sadly tame. Especially when you consider that quality is the supposed advantage of raytracing.
And c'mon. It's Gameplay that counts anyway :P
The Cell Processor
Three or four people have brought up the idea that problem would work well for the cell processor. But I don't think anyone has really seen the (rays of) light on the issue. The Cell is perfect for this. Some facts:
1) Raytracing is highly vectorized. The Cell's many processors are optimized for vector calculations.
2) Raytracing scales linearly with the number of cores. The Cell has 8 (at least in its current manifestation).
3) The Cell is already available as a PCI-Express add-in card (that even runs linux!) which sounds awfully like what a GPU is... 4) The Cell is a bitch to program. But then, so are GPUs...so maybe it's not that ridiculous to see the future of the GPU...from IBM.
How ironic it is that Intel is now pushing this technology...
It has been a nervous year, with people beginning to feel like Christian Scientists with appendicitis.
SGI had a ray tracing demo at Siggraph 2002. On the show floor, a 128-processor SGI box ran demos at around 20hz at about 512x512 pixels.. html
http://www.sci.utah.edu/stories/2002/sum_star-ray
They make some good points about geometric complexity increasing much faster than displayed pixels, so there are fewer graphics primitives per pixel, so scan-line-based algorithms will make less sense.
So in 2002 it took 128 processors to run at 20Hz at 512x512 pixels. And now we think quad-cores will be enough to render today's complex environments? That math doesn't add up to me. I think scan-line algorithms are the mainstream answer for a long time coming...
I'll let the other posters comment on the wrongness of your idea that raytracing doesn't scale with scene complexity. There was a nice SciAm article about it, if you need more convincing. Instead, I'll talk about something in the article that the other posters didn't mention. Raster Processing may scale with scene somplexity, but creation doesn't. Raster graphics must be tweaked at creation to make an object look realistic while still rendering quickly. With ray tracing, you just create an object and forget about it. It just looks right without any tweaking.
What costs game designers more: hand tweaking every object, or you buying a better computer so you can ray trace their un-tweaked objects? Now guess which way 3D graphics are gonna go...
- None can love freedom heartily, but good men; the rest love not freedom, but license. -- John Milton
There is a good article about this in August's Scientific American by W Wayt Gibbs. It's only a couple pages but worth picking up a paper issue, or if you have one of their digital subscriptions here: http://www.sciam.com/article.cfm?chanID=sa001&arti cleID=000637F9-3815-14C0-AFE483414B7F4945
E pluribus unum
I remember a talk from someone (John Carmack I think) saying something like raytracing is nice but overkill. Today's hardware maybe be able to handle realtime raytracers but no way near the quality you can get from current 3D engines.
Most special effects you see in current engines are approximation/hacks compared to what you can do with a raytracer but it's also way cheaper to compute.
It's the same kind of relationship than between texture maps vs procedural textures. Procedural textures are better for a rendering point of view, they scale better. But it's also lot harder to make a good quality texture and it requires a lot more power to render.
you are correct - pixel/fragment shading is not free - both use the same default shading model in OpenGL and probably DirectX.
the default shader is, I believe, Lambert (a close relative to Phong - if not, it's Phong) for OpenGL and probably DirectX as well. Programs in the shader can change this to whatever you want it to be (e.g. a cel shader) and you would need to do that in either a ray tracer or rasterizer.
there's a lot of things I like about ray tracing, but it's not without flaw - it handles specular highlights fantastically, but doesn't handle diffuse well at all, so you have to bolt on other techniques. Most people (including Intel) use ambient occlusion since it's a quick technique (also commonly used in polygon based graphics), but it tends to make muddy shadows (see the wikipedia entry). Radiosity is more realistic, but the patch computations are incredibly expensive (but parallel-able). photon mapping is another method that could be used, but I haven't used it myself. In college I wrote (with a team) a simple ray tracer and shortly after that class wrote a radiosity engine, so I'm familiar with both techniques. I never did really understand how to combine them, but I remember seeing POVRAY do it in the mid 90s and really wanted to figure out how they did it (but I graduated and was putting in startup hours, so that never happened).
Oh, and waves on a lake are non-trivial - to be completely realistic, you need to deal with subsurface diffusion (or estimation of), foam and caustics (if you can see through the semi-transparent water surface). The specular mirror effect would be nice, but I don't see true caustics from either a raytracer or a rasterizer (you'd need to use ray bars or cones, probably).