devtools/eigen
1
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2026-04-10 11:34:33 +08:00
Code Issues Packages Projects Releases Wiki Activity

GPU: Add library dispatch module (DeviceMatrix, cuBLAS, cuSOLVER)

Browse Source 
Add Eigen/GPU module: A standalone GPU library dispatch layer where
DeviceMatrix<Scalar> operations map 1:1 to cuBLAS/cuSOLVER calls.
CPU and GPU solvers coexist in the same binary with compatible syntax.

Core infrastructure:
- DeviceMatrix<Scalar>: RAII dense column-major GPU memory wrapper with
  async host transfer (fromHost/toHost) and CUDA event-based cross-stream
  synchronization.
- GpuContext: Unified execution context owning a CUDA stream + cuBLAS
  handle + cuSOLVER handle. Thread-local default with explicit override
  via setThreadLocal(). Stream-borrowing constructor for integration.
- DeviceBuffer: Typed RAII device allocation with move semantics.

cuBLAS dispatch (expression syntax):
- GEMM: d_C = d_A.adjoint() * d_B (cublasXgemm)
- TRSM: d_X = d_A.triangularView<Lower>().solve(d_B) (cublasXtrsm)
- SYMM/HEMM: d_C = d_A.selfadjointView<Lower>() * d_B (cublasXsymm)
- SYRK/HERK: d_C = d_A * d_A.adjoint() (cublasXsyrk)

cuSOLVER dispatch:
- GpuLLT: Cached Cholesky factorization (cusolverDnXpotrf + Xpotrs)
- GpuLU: Cached LU factorization (cusolverDnXgetrf + Xgetrs)
- Solver chaining: auto x = d_A.llt().solve(d_B)
- Solver expressions with .device(ctx) for explicit stream control.

CI: Bump CUDA container to Ubuntu 22.04 (CMake 3.22), GCC 10->11,
Clang 12->14. Bump cmake_minimum_required to 3.17 for FindCUDAToolkit.

Tests: gpu_cublas.cpp, gpu_cusolver_llt.cpp, gpu_cusolver_lu.cpp,
gpu_device_matrix.cpp, gpu_library_example.cu
Benchmarks: bench_gpu_solvers.cpp, bench_gpu_chaining.cpp,
bench_gpu_batching.cpp
This commit is contained in:
Rasmus Munk Larsen
2026-04-09 16:15:39 -07:00
parent 6a9405bf7a
commit 58c44ef36d
30 changed files with 5680 additions and 24 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
ci/test.linux.gitlab-ci.yml
View File

@@ -265,23 +265,23 @@ test:linux:x86-64:nvhpc-26.1:default:unsupported:
tags:
- saas-linux-medium-amd64-gpu-standard
# GCC-10, CUDA-12.2
test:linux:cuda-12.2:gcc-10:
# GCC-11, CUDA-12.2
test:linux:cuda-12.2:gcc-11:
extends: .test:linux:cuda
image: nvidia/cuda:12.2.0-devel-ubuntu20.04
needs: [ build:linux:cuda-12.2:gcc-10 ]
image: nvidia/cuda:12.2.0-devel-ubuntu22.04
needs: [ build:linux:cuda-12.2:gcc-11 ]
variables:
EIGEN_CI_CXX_COMPILER: g++-10
EIGEN_CI_CC_COMPILER: gcc-10
EIGEN_CI_CXX_COMPILER: g++-11
EIGEN_CI_CC_COMPILER: gcc-11
# Clang-12, CUDA-12.2
test:linux:cuda-12.2:clang-12:
# Clang-14, CUDA-12.2
test:linux:cuda-12.2:clang-14:
extends: .test:linux:cuda
image: nvidia/cuda:12.2.0-devel-ubuntu20.04
needs: [ build:linux:cuda-12.2:clang-12 ]
image: nvidia/cuda:12.2.0-devel-ubuntu22.04
needs: [ build:linux:cuda-12.2:clang-14 ]
variables:
EIGEN_CI_CXX_COMPILER: clang++-12
EIGEN_CI_CC_COMPILER: clang-12
EIGEN_CI_CXX_COMPILER: clang++-14
EIGEN_CI_CC_COMPILER: clang-14
##### arm ######################################################################