News for hackers. Stuff that matters.
Re:Run CUDA, unmodified, on AMD GPUs by amy-petrik-214 on Monday July 15, 2024 @10:50PM (#40973322)
we actually also saw this historically with openGL. openGL comes from an ancient company whispered about by the elderly programmers (30 + year old) known as SGI. Originally it was CLOSED SOURCE and SGI called it "SGI-GL" for a computer codename IRIS which was cool looking with bright popping color plastic and faux granite keyboard. Good guy SGI open sourced SGI-GL to become what we called "openGL" (get it, now it's open), and then it stuck.

That's all to say NVIDIA could pull a SGI and open their stuff, but they're going more sony style and trying to monopolize. Oh, and SGI also wrote another ancient lore library known as "STL" or the "SGI Template Library" which is like the original boost template metaprogramming granddaddy

Re:Run CUDA, unmodified, on AMD GPUs by adrian_b on Tuesday July 16, 2024 @01:25AM (#40973841)
Also the XFS file system.
Re:Run CUDA, unmodified, on AMD GPUs by usr1106 on Tuesday July 16, 2024 @02:27AM (#40974046)
Nice story, but is it correct? Wikipedia says STL was first implemented by HP and later by the same authors at SGI.
Re:Run CUDA, unmodified, on AMD GPUs by adrian_b on Tuesday July 16, 2024 @11:15AM (#40977354)
STL started even earlier, obviously without using the name "STL", as a library of generic algorithms for the programming language Ada (David R. Musser & Alexander A. Stepanov, 1987).
Re:Run CUDA, unmodified, on AMD GPUs by gpderetta on Thursday July 18, 2024 @05:08AM (#40993819)
Rumor is that STL stands for STepanov and (Meng) Lee.