Devin Matthews
Devin Matthews
@evenieri sorry for not responding to this sooner. Thanks for the contribution! Before merging we'll just want to make sure CI testing is set up, add configuration auto-detection support (if...
Hi @ivan-pi I'm going to leave this open since you're right, it is an inconsistency. This could also affect C++ users.
@fgvanzee `void*` is great for kernels since a) they're internal and b) they're typically being called through a type-erased interface anyways. For user-level typed functions, having strongly typed arguments is...
@bartoldeman I haven't done as thorough a review as @hominhquan but I especially like that you were able to find a way to convince the compiler to keep the AB...
Yes just those two settings would be a great place to start. We'd just want to give them "BLISier" names and write some documentation. We'd want to find somebody to...
A lot of that is probably just from function call overhead and a little bit of dynamic memory allocation. The AOCL version of BLIS includes a special codepath for very...
BTW I'm pretty sure that the veclib implementation doesn't switch to GPU at larger sizes but uses the Apple AMX accelerator which is on-die.
Regarding batching, if this is a common use-case then that would be a possible avenue to increase performance as doing batched gemm properly (through the existing BLAS-like APIs) would naturally...
Hi @realbabilu compiling in WSL should be very similar to compiling in linux. I would not recommend setting any machine-specific flags in CFLAGS, BLIS will take care of this internally....