Linked by Thom Holwerda on Fri 20th Jun 2008 19:58 UTC, submitted by Aron Allen
Thread beginning with comment 319470
To read all comments associated with this story, please click here.
To read all comments associated with this story, please click here.





Member since:
2007-09-28
The interesting thing about LLVM is that OpenCL probably is based on it. So, writing code for your GPU is the same as writing code for the host CPUs (maybe even in ObjC). The OS can then dispatch the code after JIT compilation to an appropriate processing unit. Maybe this is the breakthrough in multicore Apple announced. That means, they can dispatch the same code to a heterogeneous processing unit infrastructure.
Edited 2008-06-21 13:03 UTC