Slashdot Mirror

← Back to Stories (view on slashdot.org)

Audio Processing on Your Graphics Card?

Posted by michael on from the sense-substitution dept.

edsarkiss writes "BionicFX has announced Audio Video EXchange (AVEX), a technology that transforms real-time audio into video and performs audio effect processing on the GPU of your NVIDIA 3D video card, the latest of which are apparently capable of more than 40 gigaflops of processing power compared to less than 6 gigaflops on Intel and AMD CPUs." Another reader points out a story on Tom's Hardware.

16 of 335 comments (clear)

  1. Makes perfect sense... by WIAKywbfatw · · Score: 4, Interesting

    The amount of silicon on an average GPU overtook the amount of silicon on the average CPU some time ago.

    Having all that processing power available to do more than just shift pixels makes perfect sense. I'm just surprised that nobody thought of doing it sooner.

    --

    "Accept that some days you are the pigeon, and some days you are the statue." - David Brent, Wernham Hogg
    1. Re:Makes perfect sense... by hazem · · Score: 5, Interesting

      And the idea of using a specialized processor for a different purpose is not new.

      Back in 386 days, one of our professors was working on liquifaction (the ground sometimes behaves like a liquid during earthquakes). The models were very trig-intensive and took forever on a desktop. So, he wrote the simulations in Postscript and sent them to the printer where its processor could do the work much faster.

  2. Great for audio workstations... by MP3Chuck · · Score: 3, Interesting

    While most audio workstations may not have great video cards at the current time, I'd go spend $500 on a video card that'd take 90% of the workload off my processor while mixing ... it's cheaper than a lot of equipment out there.

    And the ability to get a few frags in while the band is taking a break isn't too bad either! ;)

    --
    Who doesn't like free music?
    1. Re:Great for audio workstations... by javaxman · · Score: 4, Interesting
      From the Tom's Hardware article:
      So far Cann cannot take as much performance away from the GPU as he would like. "Right now, getting the data back from the video card is very slow, so the overall performance isn't even close to the theoretical max of the card. I am hoping that the PCI Express architecture will resolve this. This will mean more instances of effects running at higher sample rates," he said.

      so it appears that there may really be a problem here... a GPU will normally do a bunch of calculations, then the raster goes *out* to the monitor, not *back* to the bus... I can see how getting data back out to the bus might be an issue. A "real" DSP/audio card would certainly be better, and they aren't *all* as expensive as the original article would have you believe... a quick google found at least one decent-looking DSP card for ~$500 out there, and I'm sure there are others, probably for cheaper ( the quoted price is for a card *and* a stack of software ), if you looked around a bit... if you're considering plunking down the cash for a PCI-X machine and a good GPU, you probably have a ~$500 for a good DSP card, too, and a special-purpose solution *designed* for the purpose at hand is almost always going to be better than repurposing a *different* special-purpose product.

      Did that make sense? What I'm trying to say is that you'd be much better off buying an actual DSP audio card than buying two GPUs. That'd just be silly. This repurposed GPU stuff is just for folks unwilling to buy an extra card, but who have a nice GPU already.

  3. mad possible by Doom by dirvish · · Score: 2, Interesting
    Brought to you thanks to Doom:
    In January 1993, John Carmack sent a press release announcing a "technical revolution in PC programming" on 386sx processors - a real-time, 256 color 3D game that let you play simultaneously with three other people. Doom was born and the desktop video game industry took off creating an impetus that pushed video and stream processors to the point they are today. Here's to you John Carmack! The repurposing of a GPU for digital audio processing would not be possible without your passion and influence.
    --
    FoundNews.com - get paid to blog.,
  4. Re:Price range of $200 to $800... by SmittyTheBold · · Score: 3, Interesting

    A GPU is much faster, but only when doing certain very specific types of operations. If you tried to write a word processor for the GeForce, it would at best run terribly slow, and at worst be an impossible task.

    GPUs are not really all that powerful compared to a CPU, but they're working with a totally different set of constraints.

    --
    ± 29 dB
  5. Coprocessor? by phorm · · Score: 2, Interesting

    Basically, from what I can glean of the article, it is basically making use of your video card's GPU as a co-processor. It doesn't state that the GPU is better at processing audio, just that in many instances it is mostly idle and thus available.

    The GPU is of course heavily optimized (over a regular CPU) for video, and perhaps some of those optimizations would be passed on to audio as well. In the future, if such things pick up, one might well see more "multimedia" card which would incorporate a mixed GPU/SPU or perhaps dual processors?

  6. like apples core image by acomj · · Score: 2, Interesting

    Apple is creating libs to work with many graphics cards acceleration for image processing. The demo was real time effects on video.

    Supports ATI and NVideo (lib figures out if you have a useable graphics card, else it just uses the cpu)
    http://www.apple.com/macosx/tiger/core.html

  7. Goes to show... by JediDan · · Score: 3, Interesting

    what dedicated hardware can do. It's an proven fact and anyone that works with embeded systems can testify to the performance. We need to stop flaunting 3+ gigahertz processors using archaic instruction sets and focus on routing data to hardware that can handle the task.

    If the CPU was nothing but a router and directed data to dedicated hardware (network cards, GPU with integrated physics engine, harddisk controller, etc) we can get away from inefficient execution tied up in an architecture that 99% of the market depends on.

    Computers were built with modularity in mind. We need to get back to those roots as it's not only a good idea, but the only way we're going to get past some performance barriers.

    --
    - Dan
  8. Apple's Core Video Technology by ZackSchil · · Score: 2, Interesting

    Apple's core image and video technology allows you to write your own processing algorithms to be run on the video card. I can imagine something like this being used to process audio in Mac OS X.

  9. Video processing -- aka MPEG2 encoding? by swb · · Score: 2, Interesting

    There's a blurb on the 6 series of GeForce cards that claim they can do video transcoding; since an hour of 2 pass encoded MPEG2 video takes my P4-3.2c about 2.5 hours, I'd love to get it at least 1x real time encoding speed (for 2-pass encodes) or at least 2x real time (for 1-pass encodes).

    Anyone know any more about this? Audio is nice, but its not nearly as CPU intensive as video transcoding.

  10. SETI/Folding by Remlik · · Score: 2, Interesting

    So can the SETI guys use spare GPU cycles? I know my work machine uses less video than CPU.

    --
    Apple free since 1990!
  11. How you can trust in graphic memory stability? by faragon · · Score: 2, Interesting

    Most graphic cards are in a big trouble on data integrity, only SRAM area is 100% sure (you'll be glad about your graphic DRAM behaviour on most cards). If you can not be 100% sure that all the bits are coherent, you'll be in a trouble if you want to do some "sensible" processing that could depend on just one bit (both raw audio and video are relatively inmune to these faults, but what about compressed/processed data to be retrieved?!).

  12. Re:Hmmm by Christ-on-a-bike · · Score: 2, Interesting
    A primates have a big-ass visual cortex. In fact, up to 50% of the brain is involved in processing visual information. It's therefore unsurprising to see such investment in video I/O for computers.

    In other words: the interfaces of a computer are (often) intended to provide immersive experiences for their users. Computer users are humans, so you would expect the processing power dedicated to each component of I/O to reflect the discernment of humans in their corresponding sense.

    In yet more words: if a dog designed a computer, it would have a crappy GPU but a beast of a smell-processing system.

  13. 40gflops?! how well does it crack dnet keys? by Narcocide · · Score: 3, Interesting

    i mean seriously... what would you ever need that much audio processing power for? distributed key cracking however....

  14. Re:Latency? by Sleen · · Score: 2, Interesting

    Great question...I am in the instrumentation business where its all about latency. Reading through the pdf, bionicFX very much claims REALTIME processing, which one may take as meaning SMALL BUFFERS...hopefully right?

    Also, "which is unfortunate as reverb is the greatest reason to use a card like this, and it is a bus effect typically, and the extra delay incurred acts to set a huge, usually inappropriate predelay."

    Which is why their first stated proof of concept algorithm will be a convolution based verb...says they don't even have to enter frequency domain.

    Regarding your protools comments: yeah! Still Envogue, and much better integration all around. But it comes down to chip cost and supply for that system to remain viable. This whole topic has arisen because of asymmetry between cpu and gpu efficiences...and most software instrument companies exist because CPU's themselves are cheaper than DSP if you count development and fabrication costs...There will always be a need for protools like integration, but given their very proprietary approach, it will suffer those characteristic inefficiencies having to do with information and standards propagation.

    Its definitely all about latency...and YES 20 ms absolutely matters!!!! Especially when you start chaining things!!! And something I always remind people: the number is not lowest latency, but lowest average ARTIFACT FREE buffer size at HIGH CPU LOADS. The only company that can reliably deliver sub 5ms latency is RME...

    Electronic Instruments Volume 2 for Reaktor on Sept. 17th. Fuckin Intense!!! Intelligent Randomization, Fast FX, and a gorgeous hybrid synth called Photone. If I could run this on an Nvidia GPU, I would pee myself because snapshot morphing is SO intensive...but sounds incredible!!!!!

    ta,

    sleen