GPGMM
GPGMM copied to clipboard
intel
GPGMM, a General-Purpose GPU Memory Management Library.
Need a call-back API so spawned threads accessed by capture tools can be exempt.
Need to figure out why x86 builders don't work with thread writer tests. Example, https://github.com/intel/GPGMM/actions/runs/3988723505/jobs/6840256571.
See https://github.com/intel/GPGMM/actions/runs/3491420677/jobs/5844031949
We need DXCore to enumerate VPU device, so should we replace IDXGIAdapter with IDCoreAdapter? wdyt? @bjjones @bbernhar @[junwei](https://github.com/fujunwei)
Examples, https://github.com/intel/GPGMM/actions/runs/2778014818 https://github.com/intel/GPGMM/actions/runs/2778014810
Adds support for a async-version of CreateResource. The idea is to allow apps that can create (but not use) resources ahead-of-time so creation doesn't bottleneck the CPU upon compute.
Replace `new` and `delete` to allocate and de-allocate, respectively, through a allocator that re-cycles CPU heaps.
Budget change events occur less frequent resulting into larger budget changes and thus more heaps to evict at once. If some new heaps must be made resident, when over budget,...