SuperPi on CUDA

[INKster]

The announcement: http://www.xtremesystems.org/forums/showthread.php?t=189034

Boy, just 30 days to port it from x86 to CUDA. Now that's fast. :wink:

[trinibwoy]

Hmmm Nvidia is really attacking on all fronts with this thing.

Folding@Home, LAME, Photoshop, PhysX, SuperPi ..... the list grows longer every day.

[INKster]

Hmmm Nvidia is really attacking on all fronts with this thing.

Folding@Home, LAME, Photoshop, PhysX, SuperPi ..... the list grows longer every day.

With all this going on, i'm beginning to wonder if a small embedded Linux kernel could be ported to CUDA in the next few months/years...
I know it's not easy, and speed would likely be an issue -even with a brute force approach-, but the mere concept is intriguing.

Another angle could be made clear by Nvidia. By placing a small general-purpose ARM CPU onboard the GPU PCB, they could have a general CELL-like scheme on a card, with the GPU acting as the SPE array.
Dreams... :razz:

[NocturnDragon]

Hmmm Nvidia is really attacking on all fronts with this thing.

Folding@Home, LAME, Photoshop, PhysX, SuperPi ..... the list grows longer every day.

It grows even bigger if you add apps that aren't CUDA accelerated...

http://blogs.adobe.com/jnack/2008/05/pixel_bender_no.html
OpenGL-accelerated Photoshop technology.;)

[Arnold Beckenbauer]

Hmmm Nvidia is really attacking on all fronts with this thing.

Folding@Home, LAME, Photoshop, PhysX, SuperPi ..... the list grows longer every day.

What? Photoshop? I don't think so: (http://blogs.adobe.com/jnack/2008/05/oct_1.html)
"Oct. 1" (aka, "Just make something up") (http://blogs.adobe.com/jnack/2008/05/oct_1.html)


It seems that news of the demo (http://blogs.adobe.com/jnack/2008/05/pixel_bender_no.html) I did the other day (a repeat of what we'd shown publicly (http://blogs.adobe.com/jnack/2008/05/technology_snea_1.html) three weeks earlier) is bouncing all around the online tech press. People are excited that the Photoshop team is exploring ways to make the app feel faster and smoother, and that's all good. What's irritating, though, is just how much bogus info is getting invented, passed around, and swallowed without question.

Gizmodo is repeating (http://gizmodo.com/393137/photoshop-cs-4-will-use-your-graphics-card-to-run-at-light-speed-do-fancy-3d-tricks) info found on a site called TG Daily (http://www.tgdaily.com/content/view/37611/140/), stating that "Photoshop CS4" (a term that I've never heard anyone from Adobe use publicly) "is expected to be released on October 1." Uhh... expected by whom? And based on what?

I didn't say anything about schedule. In fact, I never said that any of this stuff is promised to go into any particular version of Photoshop. Rather, as with previous installments (http://blogs.adobe.com/jnack/2007/10/adobe_puts_inse.html), it's a technology demonstration of some things we've got cooking--nothing more.

Doesn't matter, though: Someone pulled a date apparently out of thin air, and now everyone who can copy & paste is dutifully repeating it. The fish story grows with the telling, too. In addition to repeating the date, Electronista (http://www.electronista.com/articles/08/05/23/adobe.cs4.gpu.use/) is inventing new details (e.g. "CS3 has already had limited support for graphics processing units (GPUs) for certain filters"; sorry, no; "An upcoming wave of video cards with special physics processing will also help, Adobe explains"; nope, didn't say that; and more). Where do people get this stuff? It's particularly annoying to see made-up info presented as a response from Adobe--to questions that were never asked. (Contacting Adobe PR, or me directly, to confirm some detail isn't exactly tough.)

I'm not feeling a lot of confidence in the tech press these days. People just make up whatever they want (http://blogs.zdnet.com/microsoft/?p=1367), creating a bunch of expectations & misperceptions that people like me have to try to unravel. There's no disincentive to doing so: the sites still get their ad impressions, and clearly bloggers and readers are all too happy to take what they read at face value.

I don't know what to tell you, as the quest for ad bucks is eroding journalistic standards across the board. "Caveat lector (http://en.wikipedia.org/wiki/Caveat_lector)," and I'll keep trying to share actually legitimate information here

[trinibwoy]

Haha easy guys. Ok, remove Photoshop from the list and the hounds can lie.

[Davros]

i wounder if one day cuda could do 3d audio for example wave tracing and how about occlusion
for example if you rendered the scene to texture looking at the sound source (just the z values) you could work out occlusion for the primary wave from that

[Karoshi]

i wounder if one day cuda could do 3d audio for example wave tracing and how about occlusion
for example if you rendered the scene to texture looking at the sound source (just the z values) you could work out occlusion for the primary wave from that

Add 1 channel colour for intensity and a shadow map and you could do a nice multipath integral on the sound. But then it's not cuda, just normal RTT, with a suitable 'right enough' texture size (256x256 MSAA?). Maybe use CUDA for the integration step. And anyway, wouldnt this be overkill^3?
<duck> And once on the exp(overkill) path, wouldnt RT be the "rightier" choice for a higher number of reflections? </duck>

[Davros]

possibly karoshi but look at all the stuff the x-fi does (10,000 MIPS) + plus there is a chance high end sound cards are dying, with most people thinking ac97 codec is enough (hopefully openal can re-ignite interest in 3d audio, but im not holding my breath)

ps: google is damm quick i typed multipath integral into google and the above post came up only minutes after it was posted)

[clamb]

It grows even bigger if you add apps that aren't CUDA accelerated...

http://blogs.adobe.com/jnack/2008/05/pixel_bender_no.html
;)

Well, it's not quite the same but 'CUDA Photoshop Plug-ins' halfway down the page.

http://www.nvidia.com/object/cuda_get.html