You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
However, rather than creating multiple targets to support compiling Emu code to, I would prefer to compile Emu to some sort of bytecode or S-expressions and have a trait that can be implemented to generate CUDA or OpenCL or maybe even something else.
Note that if you want to run Emu functions on NVIDIA GPUs, many - if not all - are capable of having compiled OpenCL code run on them.
Hey
is it planned to support binding to CUDA in the future?
It would be a great enhancement I think :)
Best regards
Chips
The text was updated successfully, but these errors were encountered: