From the article: "NVIDIA awakened the world to computer graphics when it invented the GPU in 1999."
Umm....
3dfx: ookay sadface
I'm guessing they're not counting 3D chips without hardware transform & lighting support (where the 1999 GeForce 256 was the first) as "GPUs". Also, 3Dfx got taken over by nVidia soon after, so I guess it's not completely wrong. Also... press releases aren't exactly the best place to look for unbiased information.
Yeah, I was a Gamer back then and the Voodoo series were called "3D Accelerators". NVidia coined the term "GPU" for the first GeForce.
The front page is constantly changing, I didn't see it 4 days ago, and apparently a lot of other people didn't either. If you don't like it, just ignore it.
This is awesome, but I don't quite understand how it works. What kind of code will you write to invoke massively parallel behavior in LLVM? Will every language have `kernel<<<...>>>`?
Nice, but I've always though it would be nice to have a slightly higher level virtual machine that had more explicit mechanisms for expressing parrallelism.