Searched refs:gpu (Results 1 – 17 of 17) sorted by relevance
167 | `/gpu/cuda/ref` | Reference pure CUDA kernels | Yes …168 | `/gpu/cuda/shared` | Optimized pure CUDA kernels using shared memory | Yes …169 | `/gpu/cuda/gen` | Optimized pure CUDA kernels using code generation | No …172 | `/gpu/hip/ref` | Reference pure HIP kernels | Yes …173 | `/gpu/hip/shared` | Optimized pure HIP kernels using shared memory | Yes …174 | `/gpu/hip/gen` | Optimized pure HIP kernels using code generation | No …177 | `/gpu/sycl/ref` | Reference pure SYCL kernels | Yes …178 | `/gpu/sycl/shared` | Optimized pure SYCL kernels using shared memory | Yes …181 | `/gpu/cuda/magma` | CUDA MAGMA kernels | No …182 | `/gpu/cuda/magma/det` | CUDA MAGMA kernels | Yes …[all …]
544 CUDA_BACKENDS = /gpu/cuda/ref /gpu/cuda/shared /gpu/cuda/gen560 HIP_BACKENDS = /gpu/hip/ref /gpu/hip/shared /gpu/hip/gen576 SYCL_BACKENDS = /gpu/sycl/ref /gpu/sycl/shared /gpu/sycl/gen603 MAGMA_BACKENDS = /gpu/cuda/magma /gpu/cuda/magma/det617 MAGMA_BACKENDS = /gpu/hip/magma /gpu/hip/magma/det
18 python generate_tuning.py -arch a100 -max-nb 32 -build-cmd "make" -ceed "/gpu/cuda/magma"25 specifies the backend to use, typically one of `/gpu/cuda/magma` or26 `/gpu/hip/magma`.33 ./tuning "/gpu/cuda/magma"
29 - Added non-tensor basis support to code generation backends `/gpu/cuda/gen` and `/gpu/hip/gen`.30 - Added support to code generation backends `/gpu/cuda/gen` and `/gpu/hip/gen` for operators with b…57 - Added Sycl backends `/gpu/sycl/ref`, `/gpu/sycl/shared`, and `/gpu/sycl/gen`.133 - Refactored `/gpu/cuda/shared` and `/gpu/cuda/gen` as well as `/gpu/hip/shared` and `/gpu/hip/gen`…134 - Enabled support for `p > 8` for `/gpu/*/shared` backends.148 …nsor basis kernels (and element restriction kernels, in non-deterministic `/gpu/*/magma` backends).242 - New HIP MAGMA backends for hipMAGMA library users: `/gpu/hip/magma` and `/gpu/hip/magma/det`.243 - New HIP backends for improved tensor basis performance: `/gpu/hip/shared` and `/gpu/hip/gen`.273 - New HIP backend: `/gpu/hip/ref`.287 - The `/gpu/cuda/reg` backend has been removed, with its core features moved into `/gpu/cuda/ref` a…[all …]
13 gpu
39 …For example, `/gpu/cuda/gen` falls back to `/gpu/cuda/ref` for missing {ref}`CeedOperator` precond…40 …implemented method is called, then the parent `/gpu/cuda/gen` {ref}`Ceed` object uses its fallback…72 The `/gpu/*/ref` backends provide basic functionality.77 The `/gpu/*/shared` backends delegate to the corresponding `/gpu/*/ref` backends.79 All other libCEED objects are delegated to `/gpu/*/ref`.82 The `/gpu/*/gen` backends delegate to the corresponding `/gpu/*/shared` backends.86 The `/gpu/*/magma` backends delegate to the corresponding `/gpu/cuda/ref` and `/gpu/hip/ref` backen…87 …e for {ref}`CeedBasis` kernels but do not have the improvements from the `/gpu/*/gen` backends for…
16 …supported by C99 and all targeted backends (i.e. CUDA for `/gpu/cuda`, OpenCL/SYCL for `/gpu/sycl`…
95 …EED is built with GPU support, you can specify a GPU backend, e.g., `/gpu/hip` or `/gpu/cuda/gen`."
217 "ceed_gpu = libceed.Ceed('/gpu/cuda')\n",
96 getresource(ceed) == "/gpu/cuda/gen" && error(98 "/gpu/cuda/gen is not compatible with user Q-functions defined with ",99 "libCEED.jl.\nPlease use a different backend, for example: /gpu/cuda/shared ",100 "or /gpu/cuda/ref",
107 Returns true if the given [`Ceed`](@ref) object has resource `"/gpu/cuda/*"` and false
25 device_type = sycl::info::device_type::gpu; in CeedInit_Sycl()
34 enum gpu { enum in sycl::ext::libceed::device_arch
44 if (DeviceType == sycl::info::device_type::gpu) { in prepareOclocArgs()
392 url={https://github.com/karlrupp/cpu-gpu-mic-comparison},
165 If MFEM is built with `libCEED` support, existing MFEM users can pass `-d ceed-cuda:/gpu/cuda/gen` …