@ Entroper

With all due respect are you "just " a PHP and SQL programmer? Do you have any experience with designing game engines, low-level development in assembler and/or C, have you hit the hardware like demoscene coders on older systems do?

I think there's a great difference with regard to perspectives from high level developers (fully dependent on what the low-level developers provide them with) and those who work on lower level stuff like writing OS kernels and programming languages like many people I know have done. The latter usually have a much deeper understanding of hardware architectures and their potential.

Of course I do value your input for this thread if you refrain from off topic personal orientated stuff. Just would like to know for sure what you mean exactly by stating "I'm a programmer"?

Now we're having a conversation :)

By "I am a programmer" I mean I've been programming since 1996, when I was a freshman in high school.  I graduated August of this year with a Bachelor's Degree in Computer Science from Virginia Tech, and in my last semester there, I actually wrote a matrix multiply for the Cell using low-level optimization techniques like pipeline optimization, use of SIMD instructions, double buffering DMA transfers, etc.  I've also worked with computer graphics for several years; my last two projects were a raytracer and a stereoscopic renderer for a space combat sim (think Tie Fighter with 3D glasses).  I suppose the closest I've come to writing commercial game code was as lead programmer for Serious Fortress, which was a TF mod for Serious Sam (couldn't get any modelers, though).

I can do PHP and SQL too.  :)  Actually all the employers here are doing .NET and SQL server, which blows, because it's not what I'm interested in.  Games and programming are my passion, so it's likely I'll end up a game developer before too long.  I'm currently working on something to rip the data and graphics out of a NES ROM, which will then be used to play the game in an updated OpenGL engine.

@ Entroper

Interesting, so what were your findings with regard to this research?



So what do you think about Cell's use for things like raytracing?

"We have shown how to efficiently map the ray tracing algorithm
to the Cell processor, with the result that a single SPE achieves
roughly the same traversal performance as the fastest known x86-
based systems, and using all of a Cell’s SPEs yields nearly an order
of magnitude higher traversal performance than on an Opteron."

http://graphics.cs.uni-sb.de/~benthin/cellrt06.pdf
http://www.youtube.com/watch?v=lr-R4bUZIQw

@ Entroper

Maybe I have been fighting fire with fire, but I have been following the Cell for quite a while and its potentials, everytime I made a remark with regard to the Cell I'm dissed as being just a Sony fanboy (maybe making me want to prove them wrong) and people here often make such wild statements, based on nothing actually, some honestly know nothing about. Initially I was just interested in the Cell, but I read so many other claims with regard to the PS3 I felt I had to research the validity (such as Blu-Ray speed, hardware failures, etc). I always had something against bullies, as the biggest kid in my neighbourhood I mostly kicked the bully's arse. (and PS3 fans had to endure lots of bullshit for liking the platform)

At least it seems I am now talking to someone who actually knows some stuff regarding this topic and I would love to hear your perspectives and initial experiences.

this does me thinking of :

http://vgchartz.com/forum/thread.php?id=680

and this oldtimer :

http://vgchartz.com/forum/thread.php?id=677

My findings were that I hate bus errors. :) Actually, as you may know, there was a shooting at VT in April, which cut the semester short. The Cell project was therefore optional. Myself and one other student were the only ones who asked for access to the cluster, and I don't think he ever coded anything. But because I only worked on it for about a week, I didn't finish it. I did get software pipelining and partial double buffering going, but I needed to reorganize the shared data access in order to get all 6 SPEs able to do double buffering while sharing the matrix data. I expect I would've ended up in the 50-75 GFLOPS range when all was said and done.

A raytracer is one of the things I'd like to write if I get my hands on the Cell again. It's never going to be usable in a game, unfortunately. Complex scene render times are measured in minutes on fast CPUs, and an order of magnitude faster than that is still more like 30 seconds per frame than 30 frames per second. But I would say that if Pixar doesn't have a rendering farm made of a few dozen PS3s, they're probably doing something wrong.

@ Entroper

You may find this interesting:

Running matrix multiplication of 4096x4096 matrices using 16 SPEsAggregated performance for all 16 SPEs: 406.37 GFLOPS (of 409.60 GFLOPStheoretical peak).link

 @ Entroper

One of the many things I like about the Cell is that it's very suitable for future multi-processing designs, for example currently you can already hook up additional PS3s to give you more potential performance (very cheaply):

"The video above contains some real-time ray-tracing, where three PS3's "render a model that is 75x more complex then those used in today's games."

http://www.ps3fanboy.com/2007/04/05/real-time-ray-tracing-on-the-playstation-3

A reason to own 3 PS3s!
http://www.youtube.com/watch?v=uh1jB4hVJRg

MikeB said: @ Entroper ut I would say that if Pixar doesn't have a rendering farm made of a few dozen PS3s, they're probably doing something wrong. One of the many things I like about the Cell is that it's very suitable for future multi-processing designs, for example currently you can already hook up additional PS3s to give you more potential performance (very cheaply): "The video above contains some real-time ray-tracing, where three PS3's "render a model that is 75x more complex then those used in today's games." http://www.ps3fanboy.com/2007/04/05/real-time-ray-tracing-on-the-playstation-3 A reason to own 3 PS3s! http://www.youtube.com/watch?v=uh1jB4hVJRg

 Or:

"Three UC Irvine students have won IBM's Cell Broadband Engine Professor University Challenge. Their project? Oh, hooking three PS3s together to reproduce human brain functions:

The group's findings, which were showcased at the 2007 Power Architecture Developer Conference in Austin, Texas focused on reproducing brain algorithms. Brain algorithms are intricate step-by-step problem-solving procedures performed by the brain to complete everyday functions."

I think nomatter if you're a Wii or XBox 360 fanboy, if you love tech these projects are pretty cool!

http://kotaku.com/gaming/self_aware/3-ps3s--part-of-a-human-brain-310685.php 

MikeB, have you noticed that almost half of the 51 posts in this thread are yours?