Windows 7 WARP system to allow for DirectX 10 CPU acceleration
By Donald Melanson 
posted November 28th 2008 12:34PM
posted November 28th 2008 12:34PM
We've already heard that Microsoft plans to make use of GPU acceleration in Windows 7, but it looks like the company is also going to be doing its part for the GPU-less out there, with the OS's new so-called WARP system promising to allow for DirectX 10 acceleration using nothing more than a plain old CPU. Among other things, that's apparently being done to avoid a recurrence of the Vista-capable debacle that happened last time around, when some systems that were said to be capable of running the OS were, in fact, anything but. According to Microsoft, WARP (or Windows Advanced Rasterization Platform) will work with as little as an 800MHz CPU, although it says it'll work better on multi-core processors with SSE 4.1. To really put it to the test, Microsoft apparently even went so far as to run a few Crysis benchmarks with the system, and managed to clock in a blistering 7.36 fps frame rate at 800 x 600 on a Core i7-equipped PC, which is actually slightly better than what Intel's current integrated graphics were able to eke out.
It's eke, not eek.
is this destined for the same success that met OS/2 Warp xD?
Bah gimme me mode 7 scalable textures
hrmmm mode 7
windows 7
Coincidence Or conspiricy?
There's a new "Direct2D" API in Win7 that is intended for that sort of stuff.
I thought that os/2 WARP was just a distant memory... until an ATM machine in Peru ate my cashcard this summer. I waited with crossed fingers, and then it rebooted itself. During the reboot,it announced itself to be OS/2 WARP but still didn't return my only cashcard.
The bank returned it to me 24 hours later, thanks be, my thanks also to the tourist police of Puno.
In the pantheon of names already used, this has to be at the top of the list.
All you computer old farts remember IBM's OS X 3? Code name: Warp.
Doh!
Speaking of DOH (and muscle memory) it was OS/2 3.0 - Warp.
OS/2 4.0 was Borg.
Watch for a Microsoft announcement soon to assimilate another company's code name.
This won't really work on atom that well, will it?
@Engadget editors
Can you possibly get some writers here that know what the hell they are talking about. Just as I mentioned in the engadget article that you linked to above referring to GPGPU/Stream processing in Windows 7, you guys continue to confuse the concept of conventional GPU acceleration of 3D graphics (e.g OpenGL applications, Direct3D games, Aero3D in Vista, Quartz in OSX, Compiz in Linux) with the concept of "GPGPU" aka general purpose processing on a GPU. GPGPU processing, also known as "stream processing", allows processor intensive applications to offload a portion of their computation onto a GPU, which acts as a highly-parallel co-processor.
Both Nvidia (CUDA) and ATI (Brook+) have in-house APIs for this type of application, but the future of GPGPU is headed towards standardized APIs that can utilize hardware from either company. Both Microsoft and Apple have initiatives for this. Apple, along with nVidia and ATI initiated the "OpenCL" specification which is housed under the Chronos Group (the organization that maintains the OpenGL spec). The first implementation of OpenCL will be in Apple's OSX 10.6 Snow Leopard.
Microsoft on the other hand is implementing a GPGPU API as part of DirectX 11 in a component called "computer shaders", which will be a new shader type in addition to the conventional pixel, vertex, and geometry shaders of Direct3D/OpenGL.
To summarize, you guys need to get all your editors on the same page and STOP referring to GPGPU processing as simply "GPU acceleration". Those are NOT the same concepts and you are confusing the hell out of the less knowledgeable among your readers.
Pretty much...
Good informative post thank
http://www.zahipedia.com/2008/12/01/microsoft-warps-cpus-to-gpus/