News

MAGMA 1.6.2 released
2015-05-05
MAGMA 1.6.2 is now available. This is primarily a bug fix for MAGMA sparse.
  • Added magma_{s,d,c,z}sqrt for real and complex scalar square root.
  • Added magma_ceildiv and magma_roundup.
  • Fixed magmablas_zlaset and magmablas_zlacpy for large M or N > 4M.
  • Fixed testers for geqrf_batched and trsm_batched to compile with CUDA 5.x.
MAGMA sparse
  • All allocation failures and other errors now return error codes.
  • cuSPARSE error codes mapped to MAGMA error codes.
  • LOBPCG sparse eigensolver enabled for preconditioning using Jacobi and incomplete LU factorizations.
  • Some name changes in MAGMA sparse for consistency with dense MAGMA.
  • Added a tester for the sparse level 1 BLAS.
  • Bug fix in complex FGMRES.
  • Added iterative incomplete factorization routines (iterative ILU/iterative IC).
  • Enhance the ILU/IC with fill-in (level-ILU).

See the Downloads section for a download link.


MAGMA 1.6.1 released
2015-01-30
MAGMA 1.6.1 is now available. This release provides performance improvements and increased functionality. More information is given in the MAGMA 1.6 Quick Reference Flier. The MAGMA 1.6.1 release adds the following new functionalities:
  • Building as both shared and static library is default now.
    Comment out FPIC in make.inc to build only static library
  • Added max norm and one norm to [zcsd]lange
  • Extended {sy|he}mv and {sy|he}mv_mgpu implementation to upper triangular
  • Fixed memory access bug in {sy|he}mv_mgpu, used in {sy|he}trd_mgpu
  • Fixed errant argument check in laswp, affecting getrf_mgpu
  • Fixed tau in [cz]gelqf, which needed to be conjugated
  • Fixed workspace size in symmetric/Hermitian eigenvalue solvers
  • Made fast magmablas_zhemv default in symmetric/Hermitian eigenvalue solvers
    (previously needed to define -DFAST_HEMV option)
  • Added FGMRES for non-constant preconditioner operator
  • Added backward communication interfaces for SpMV and preconditioner passing the vectors on the GPU
  • Added function to generate cuSPARSE ILU level-scheduling information for a given matrix
  • Adding the batched QR routine
  • Performance improvments of all batched routines
  • Fixing "nan" output for batched factorizations.

Support for the new Tesla K80 "GK210-Duo" is provided through MAGMA's multiGPU routines (see the MAGMA LU Benchmark on up to four K80s).

See the Downloads section for a download link.


MAGMA MIC 1.3.1 for Intel Xeon Phi Coprocessors Released
2015-01-30

MAGMA MIC 1.3.1 is now available. This release provides implementations for MAGMA's one-sided (LU, QR, and Cholesky) and two-sided (Hessenberg, bi- and tridiagonal reductions) dense matrix factorizations, as well as linear and eigenproblem solver for Intel Xeon Phi Coprocessors. More information on the approach is given in this presentation.

The MAGMA MIC 1.3.1 release adds
  • Added orthogonal transformations routines
    {zun|cun|dor|sor}mbr 
    {zun|cun|dor|sor}mlq
    {zun|cun|dor|sor}mql
  • Added SVD routine using divide and conquer algorithm
    {z|c|d|s}gesdd
  • Performance optimizations for the two-sided factorizations
    (reductions to bidiagonal, tridiagonal, and upper Hessenberg)
  • Added zscal, hemv for CPU, copy functions
  • Added LDLt without pivoting
  • Added hybrid solver for symmetric indefinite problems using the Bunch-Kaufman diagonal pivoting method
    {zhe|che|dsy|ssy}sv

See the Software section for a download link.


MAGMA 1.6.0 released
2014-11-15
MAGMA 1.6.0 is now available. This release provides performance improvements and increased functionality. More information is given in the MAGMA 1.6 Quick Reference Flier. The MAGMA 1.6.0 release adds the following new functionalities:
  • MAGMA Batched linear algebra routines:
    • BATCHED MAGMA BLAS including gemm, gemv, herk, and trsm
    • BATCHED LU, GETRI, and Cholesky factorizations
  • Bunch-Kaufman factorization (and solver) for symmetric indefinite matrices
    {z|c|d|s}{he|sy}trf
  • Non-pivoted LDLt
  • Random Butterfly Transformation (RBT) and solver based on RBT + LU without pivoting + iterative refinement
  • MAGMA Sparse routines.

There are also improved testers and bug fixes. Support for the new Tesla K80 "GK210-Duo" is provided through MAGMA's multiGPU routines (see the MAGMA LU Benchmark on up to four K80s).

See the Downloads section for a download link.


MAGMA MIC 1.3 for Intel Xeon Phi Coprocessors Released
2014-11-15

MAGMA MIC 1.3 is now available. This release provides implementations for MAGMA's one-sided (LU, QR, and Cholesky) and two-sided (Hessenberg, bi- and tridiagonal reductions) dense matrix factorizations, as well as linear and eigenproblem solver for Intel Xeon Phi Coprocessors. More information on the approach is given in this presentation.

The MAGMA MIC 1.3 release adds
  • Usage and performance improvements
  • Bunch-Kaufman factorization for symmetric indefinite matrices
    {z|c|d|s}{he|sy}trf
  • LU without pivoting in CPU and MIC interfaces
    {z|c|d|s}getrf_nopiv[mic]
  • Random Butterfly Transformation (RBT)
  • A new solver based on RBT + LU without pivoting + iterative refinement

See the Software section for a download link.


Displaying 1-5 of 49 Entries
Jun 29 2015 Admin Login