Industry Group to Establish OpenCL Standard

[MacRumors]

The Khronos Group, a member-funded consortium focused on establishing open standard application programming interfaces (APIs), has announced the formation of the "Compute Working Group" to focus on open standards for parallel computing across graphics processing units (GPUs) and CPUs. Apple, AMD (ATI) and Nvidia are amongst the initial members.

The group will specifically evaluate and establish Apple's proposed Open Computing Language (OpenCL) specification. OpenCL aims "to enable any application to tap into the vast gigaflops of GPU and CPU resources through an approachable C-based language."

A widely available open-standard compute programming specification with high-performance, general computation support and robust numerics will complement existing solutions and further liberate GPU-based compute power from the realm of graphics-only applications and provide a multi-vendor, portable interface for coordinating all the many-core GPUs and multi-core CPUs within a system. Such capability will have broad applicability - including a central role in the Khronos API ecosystem by providing a powerful compute front-end to OpenGL and OpenGL ES, and a platform for accelerating tasks such as physics and image processing / recognition.

The Khronos Group is also responsible for OpenGL and OpenGL ES standards as well as many others.

At WWDC, Apple first announced its plans to introduce performance enhancing technologies into their next version of Mac OS X (Snow Leopard). The technologies included "Grand Central" and "OpenCL" which promise to improve computer performance by taking advantage of modern multi-core processors as well as the GPUs found on modern video cards.

According to the president of the Khronos group, this technology could be used in both desktop and handheld devices in the future.

Article Link

[gnasher729]

With the kind of performance potential that using the GPU can give you, there will be lots of customers in need of serious floating point performance who don't care or don't care very much what operating system a computer has, only whether it supports OpenCL or not. If that works out, then Apple will sell an awful lot of MacPro's and AMD / NVidia will sell an awful lot of GPUs going into these MacPro's.

[SmurfBoxMasta]

MS....
Standards ......

Shame on ya'll fer puttin those 2 words in the same sentence

me thinks the only standards they have managed to conform to is bloatware, discombobulation, and greedscum.......

But great news for everyone else !

[137489]

Originally Posted by SmurfBoxMasta

MS....
Standards ......

Shame on ya'll fer puttin those 2 words in the same sentence

me thinks the only standards they have managed to conform to is bloatware, discombobulation, and greedscum.......

But great news for everyone else !

Yep, I remember about 7 years back I went to a conference on developing websites (mostly with client side computing - although they touched a bit on srever side). There they demo's a bunch of the newest technology (dreamweaver, MS Front Page, Java Script, and a few others).

They showed us all the HTML code that was created by these apps, then for 15 min they went over the code. They pointed out that while (at the time) front page was the easiest to use to get websites up quickly - the code was not standard. they also said to stay away from DHTML, which was created by MS. The instructures put every bell and whistle that front page offered and created a beautiful site in minutes - then demontrstated how 90% of the webrowsers at that time crashed trying to run it. Even Internet Explorer had some trouble at times. the instructor then brought up the W3C consortium that sets the standards for the internet - half if not more of the code that front page developed was not standard.

It is still even true to this day with any of their website creation software. that is why in web essentials and silver light, they took away alot of what we loved (easy navagation menues, blogs [kinda blew my mind - the web these days is 80% blogs)], guest books, proto albums, slide shows)

[Rocketman]

The first link seems broken.

Try http://www.khronos.org/

Rocketman

[Marx55]

Check out:

Intel Recommends Developers Plan for Massive Multi-Core Processing
http://www.macrumors.com/2008/07/06/...ore-processing

[Zwhaler]

Let's hope that this means that the performance improvements will justify the likely $130 upgrade.

[Peel]

Nice to see ATI and Nvidia jump in early with their support. Of course, if this pans out and gets adopted as a standard they both stand to sell lots of higher end video cards as computing power enhancements, instead of just to gamers.

[furious]

waits for microsoft to jump on board.

[fuziwuzi]

the current high end video cards draw too much power, upto 235w for the new nvidia gt200 gpus.

[ffakr]

Originally Posted by fuziwuzi

the current high end video cards draw too much power, upto 235w for the new nvidia gt200 gpus.

A high end GPU may require 3x the power (wattage) of a mid-high end cpu but a GPU has 10-20x the performance of a mid-range multi-purpose CPU.
Video cards today can push 500GigFlops of single precision FP work.
A very high end Quad Core Xeon 3GHz chip benched at 81GFlops in Linpack:
http://www.intel.com/performance/server/xeon/hpcapp.htm
I've got the 2.8s in my XServe. They're fast.
Sure GPUs run hot.. but they don't as hot as 6+ 3Gig Xeons (a 120W part).

That all said..
GPUs do a few things REALLY well. Some things they simply don't do at all.
Right now, not all of the GPUs (most) don't support double-precision floating point. They might not support the same IEEE standards for FP either. it'd probably be bad if your CPU and your GPU rounded FP numbers differently when you use them for parts of the same calculation. ;-)
GPU streams also don't do a good job talking to each other so there are issues with scheduling.
I'm glad to hear that ATI is on board already. They've got double-precision FP support in some GPUs already. They even sell a computation card that uses GPUs (though it doesn't appear to be for sale anywhere). http://ati.amd.com/products/streamprocessor/specs.html

As I understand it, OpenCL will be an abstraction layer for programmable GPUs. There should be no reason why it couldn't also support other math processors, like gaming Physics chips. The big vendor (in that small market) was recently purchased by Nvidia. :-) Physics chips are designed for a different set of tasks, and they run cooler. I'd love to see a card with a mess-o physics chips on it.
http://www.nvidia.com/object/nvidia_physx.html

GPUs are essentially scores, soon to be hundreds, of relatively discrete processing units. The current high-end GPUs have over a hundred "streams".
These are designed to run in parallel and therefore they can get a phenomenal amount of work done in a given time.
If you task can be broken up into lots of little pieces AND if your task requires an acceptable set of math functions, it'll scream on your OpenCL system. You can literally get the performance of a half-rack of cluster from your desktop [likely more due to lower latency].. certainly more if you put in a several video cards.

Most common applications won't see a lick of OpenCL acceleration. We should (hopefully) see some acceleration in apps that apply filters to digital streams. I could potentially see GarageBand, iPhoto, and iMovie running like 'a bat out of hell'. They're kind of made for this type of thing.
Where OpenCL should really pay off is in custom apps, like Computational Fluid Dynamics programs that someone like NASA might use to design a replacement for the Space Shuttle. OpenCL will probably be a huge boon for tasks like protein folding. Professional 3d rendering will likely get much faster.

None of this is new stuff mind you. A lot of people have been doing GPU programming for a long time ( I know some of them ). There are commercial apps out that do GPU programming already (not MS Office and the like.. but usually professional apps in Vertical markets like medical imaging and such).
OpenCL is cool because it's GPU programming for the masses. It's really the first time consumers can see this kind of benefit because it'll be much easier .. Way easier for developers to wrap OpenCL acceleration into Apps. In particular, Apple technologies will be OpenCL accelerated. If you embed something like Quartz Composer or if you leverage Quartz Extreme in some way.. it should just be faster. (this all depends on which Apple technologies lend themselves to OpenCL.. I'm not saying either of the previous will be accelerated but they're good candidates) :-)

[bigwig]

Originally Posted by furious

waits for microsoft to jump on board.

Microsoft generally only joins standards groups in order to coopt and destroy from within.

[CrackedButter]

Originally Posted by bigwig

Microsoft generally only joins standards groups in order to coopt and destroy from within.

Lets hope they do their own standard and not do it properly and then the performance factor between the 2 different OS's will increase due to MS's arrogance.

[furious]

Originally Posted by bigwig

Microsoft generally only joins standards groups in order to coopt and destroy from within.

kinda my point.

[DMann]

Originally Posted by Peel

Nice to see ATI and Nvidia jump in early with their support. Of course, if this pans out and gets adopted as a standard they both stand to sell lots of higher end video cards as computing power enhancements, instead of just to gamers.

Everyone stands to benefit with Open CL development; Apple, Intel, ATI, NVidia, and most significantly, the consumer and pro given the greatly enhanced boost in productivity. It is wonderful to witness the evolution of core processing, and greater still to be able to actually utilize all cores to their full potential with all applications and processes. This alone would justify a $130 upgrade, as would the integration of a bootable read/write ZFS. Already looking forward to MW 2009, and all of the iPhone apps which will be introduced along the way. Snow Leopard seems to be becoming a major upgrade to an already solid OS, as Open CL further evokes collaboration from all ends toward the common goals of efficiency, enhanced performance, and power.

[arn]

http://online.wsj.com/article/SB1213...2_1571_leftbox

Two rival chip makers are about to deliver the next advance in technology to improve the realism of videogames. But this time their efforts could have a broader impact.

AMD seems on board

The GPU makers are addressing the software challenge. Nvidia has an internally developed programming scheme called CUDA; AMD plans to use a programming technology called OpenCL that Apple Inc. and other companies are backing.

[arn]

A developer on Snow Leopard

http://macdaddyworld.com/2008/06/15/wwdc-the-line/

The things I saw this week at WWDC made my eyes bug out. They’re determined to squeeze every last cycle out of the CPU (and the GPU for that matter), constantly asking the question “How can we make this faster?”.

....

Until now, efficiently and properly taking advantage of all available cores has been a tricky and error-prone process even for the most brilliant of engineers. Snow Leopard will solve this problem in many ways, with new language features, and a new operation paradigm which shifts the burden of threaded programming away from the developer and into the capable hands of the OS.

[philips]

Originally Posted by Peel

Nice to see ATI and Nvidia jump in early with their support. Of course, if this pans out and gets adopted as a standard they both stand to sell lots of higher end video cards as computing power enhancements, instead of just to gamers.

I have impression that they - ATI and nVidia - treat Apple separately. After all nobody on Windows cares about OpenGL anymore. Yes, it is supported - but M$ forced everybody into Direct X chains.

Essentially, only open source OSs and Apple are those who would see any benefits of the new standard. Windows market would likely be torn apart by nVidia and ATI proprietary solutions to only later be "put in place" by next version of Direct X.

Also to the "they both stand to sell lots of higher end video cards" statement. It is quite ambiguous at the moment. At least ATI tries clearly to find solution for low-end. nVidia never really cared about low-end - they have fat enough margin from high end market they own at the moment. Adding such feature (CUDA) to their high end offering only adds to the exclusivity of newer video cards.

[solipsism]

Originally Posted by Zwhaler

Let's hope that this means that the performance improvements will justify the likely $130 upgrade.

If it doesn't, then there is now need to buy it. If you have a MacBook and don't need/want native Exchange support in Mail.app and Leopard is stable for you, i don't see a reason to upgrade. There won't be many new end-user features—adn the main ones may find their way into Leopard—and the OpenCL and Grand Central won't do much for a dual core system with an integrated graphics card and no more than 4GB RAM.