GitLab

* add triton ptxas path  for gb300 to find where it is to avoid compilation errors
Signed-off-by: tdophung <tdophung@nvidia.com>

* add these flags in advance to preven future breaks when ops are extended to multi gpus
Signed-off-by: tdophung <tdophung@nvidia.com>

* add this also to L1
Signed-off-by: tdophung <tdophung@nvidia.com>

---------
Signed-off-by: tdophung <tdophung@nvidia.com>