Re:The history of the modern graphics processor
by greggman on Wednesday May 28, 2014 @11:34AM
(#7810769)
I could be wrong but didn't Evans and Sutherland have 3D accelerated graphics in the 70s for military flight simulators? Or is my memory bad?
Re:The history of the modern graphics processor
by rayiner on Wednesday May 28, 2014 @01:48PM
(#7811737)
My first inclination was to start with E&S, but my understanding is that their 3D technology in the 1970's was based on vector displays: http://archive.computerhistory.org/resources/text/Evans_Suth.... Which is clearly hardware-accelerated 3D, but not quite how a modern GPU works because it doesn't involve rasterization of triangles into a framebuffer. Of course this was before my time, so I'd be happy to be corrected...
Re:The history of the modern graphics processor
by adventured on Wednesday May 28, 2014 @10:01AM
(#7809908)
Article has tons of holes in it. They didn't even mention Pixar once:
What? No.
The evolution of the modern graphics processor begins with the development of commercial 3D systems in the 1980's. Jim Clark founded SGI in 1982 based on his work into hardware acceleration of geometry computations for 3D at Stanford. By the mid-1980's, SGI workstations were able to handle 3D modeling and animation locally: http://en.wikipedia.org/wiki/Silicon_Graphics#IRIS_2000_and_....
You can find some of Clark's papers on the hardware here: http://www.computer.org/csdl/mags/co/1980/07/01653711.pdf (1980); http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.359.... (1982). I can't find a good reference describing the rasterization side of the pipeline, though the second paper describes it very briefly on page 132.
There's a really neat website going into details of early 3D consumer chips: http://vintage3d.org. It's very interesting, though again it should be noted that these products came out in the 1990's, well after hardware acceleration of 3D was well-established in the workstation market.