Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add a method cudss_set to update the sparse matrix in a CudssSolver #30

Merged
merged 2 commits into from
Mar 28, 2024
Merged

Add a method cudss_set to update the sparse matrix in a CudssSolver #30

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
32787ff
Add a method cudss_set to update the sparse matrix in a CudssSolver
amontoison Mar 28, 2024
b52063a
Update test_cudss.jl to test in-place factorizations
amontoison Mar 28, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
44 changes: 44 additions & 0 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -54,6 +54,20 @@ cudss("solve", solver, x_gpu, b_gpu)

r_gpu = b_gpu - A_gpu * x_gpu
norm(r_gpu)

# In-place LU
d_gpu = rand(T, n) |> CuVector
A_gpu = A_gpu + Diagonal(d_gpu)
cudss_set(solver, A_gpu)

c_cpu = rand(T, n)
c_gpu = CuVector(c_cpu)

cudss("factorization", solver, x_gpu, c_gpu)
cudss("solve", solver, x_gpu, c_gpu)

r_gpu = c_gpu - A_gpu * x_gpu
norm(r_gpu)
```

### Example 2: Sparse symmetric linear system with multiple right-hand sides
Expand All @@ -64,6 +78,7 @@ using CUDSS
using SparseArrays, LinearAlgebra

T = Float64
R = real(T)
n = 100
p = 5
A_cpu = sprand(T, n, n, 0.05) + I
Expand All @@ -84,6 +99,20 @@ cudss("solve", solver, X_gpu, B_gpu)

R_gpu = B_gpu - CuSparseMatrixCSR(A_cpu) * X_gpu
norm(R_gpu)

# In-place LDLᵀ
d_gpu = rand(R, n) |> CuVector
A_gpu = A_gpu + Diagonal(d_gpu)
cudss_set(solver, A_gpu)

C_cpu = rand(T, n, p)
C_gpu = CuMatrix(C_cpu)

cudss("factorization", solver, X_gpu, C_gpu)
cudss("solve", solver, X_gpu, C_gpu)

R_gpu = C_gpu - ( CuSparseMatrixCSR(A_cpu) + Diagonal(d_gpu) ) * X_gpu
norm(R_gpu)
```

### Example 3: Sparse hermitian positive definite linear system with multiple right-hand sides
Expand All @@ -94,6 +123,7 @@ using CUDSS
using SparseArrays, LinearAlgebra

T = ComplexF64
R = real(T)
n = 100
p = 5
A_cpu = sprand(T, n, n, 0.01)
Expand All @@ -114,4 +144,18 @@ cudss("solve", solver, X_gpu, B_gpu)

R_gpu = B_gpu - CuSparseMatrixCSR(A_cpu) * X_gpu
norm(R_gpu)

# In-place LLᴴ
d_gpu = rand(R, n) |> CuVector
A_gpu = A_gpu + Diagonal(d_gpu)
cudss_set(solver, A_gpu)

C_cpu = rand(T, n, p)
C_gpu = CuMatrix(C_cpu)

cudss("factorization", solver, X_gpu, C_gpu)
cudss("solve", solver, X_gpu, C_gpu)

R_gpu = C_gpu - ( CuSparseMatrixCSR(A_cpu) + Diagonal(d_gpu) ) * X_gpu
norm(R_gpu)
```
5 changes: 5 additions & 0 deletions src/interfaces.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,6 +48,7 @@ end
cudss_set(matrix::CudssMatrix{T}, v::CuVector{T})
cudss_set(matrix::CudssMatrix{T}, A::CuMatrix{T})
cudss_set(matrix::CudssMatrix{T}, A::CuSparseMatrixCSR{T,Cint})
cudss_set(solver::CudssSolver{T}, A::CuSparseMatrixCSR{T,Cint})
cudss_set(solver::CudssSolver, parameter::String, value)
cudss_set(config::CudssConfig, parameter::String, value)
cudss_set(data::CudssData, parameter::String, value)
Expand Down Expand Up @@ -84,6 +85,10 @@ function cudss_set(matrix::CudssMatrix{T}, A::CuSparseMatrixCSR{T,Cint}) where T
cudssMatrixSetCsrPointers(matrix, A.rowPtr, CU_NULL, A.colVal, A.nzVal)
end

function cudss_set(solver::CudssSolver{T}, A::CuSparseMatrixCSR{T,Cint}) where T <: BlasFloat
cudss_set(solver.matrix, A)
end

function cudss_set(solver::CudssSolver, parameter::String, value)
if parameter ∈ CUDSS_CONFIG_PARAMETERS
cudss_set(solver.config, parameter, value)
Expand Down
42 changes: 42 additions & 0 deletions test/test_cudss.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -140,6 +140,20 @@ function cudss_execution()

r_gpu = b_gpu - A_gpu * x_gpu
@test norm(r_gpu) ≤ √eps(R)

# In-place LU
d_gpu = rand(T, n) |> CuVector
A_gpu = A_gpu + Diagonal(d_gpu)
cudss_set(solver, A_gpu)

c_cpu = rand(T, n)
c_gpu = CuVector(c_cpu)

cudss("factorization", solver, x_gpu, c_gpu)
cudss("solve", solver, x_gpu, c_gpu)

r_gpu = c_gpu - A_gpu * x_gpu
@test norm(r_gpu) ≤ √eps(R)
end
end

Expand Down Expand Up @@ -171,6 +185,20 @@ function cudss_execution()

R_gpu = B_gpu - CuSparseMatrixCSR(A_cpu) * X_gpu
@test norm(R_gpu) ≤ √eps(R)

# In-place LDLᵀ / LDLᴴ
d_gpu = rand(R, n) |> CuVector
A_gpu = A_gpu + Diagonal(d_gpu)
cudss_set(solver, A_gpu)

C_cpu = rand(T, n, p)
C_gpu = CuMatrix(C_cpu)

cudss("factorization", solver, X_gpu, C_gpu)
cudss("solve", solver, X_gpu, C_gpu)

R_gpu = C_gpu - ( CuSparseMatrixCSR(A_cpu) + Diagonal(d_gpu) ) * X_gpu
@test norm(R_gpu) ≤ √eps(R)
end
end
end
Expand Down Expand Up @@ -202,6 +230,20 @@ function cudss_execution()

R_gpu = B_gpu - CuSparseMatrixCSR(A_cpu) * X_gpu
@test norm(R_gpu) ≤ √eps(R)

# In-place LLᵀ / LLᴴ
d_gpu = rand(R, n) |> CuVector
A_gpu = A_gpu + Diagonal(d_gpu)
cudss_set(solver, A_gpu)

C_cpu = rand(T, n, p)
C_gpu = CuMatrix(C_cpu)

cudss("factorization", solver, X_gpu, C_gpu)
cudss("solve", solver, X_gpu, C_gpu)

R_gpu = C_gpu - ( CuSparseMatrixCSR(A_cpu) + Diagonal(d_gpu) ) * X_gpu
@test norm(R_gpu) ≤ √eps(R)
end
end
end
Expand Down
Loading