OpenBLAS

OpenBLAS copied to clipboard

DGESVD slow coparing with intel implementation

49

Dears, when I run this command on a scilab running on freebsd compiled with openblas I get a time of 37 seconds A1=rand(1000, 1000); tic();[S, v, D]=svd(A1);toc() ans = 37.276...

OtacilioNeto

cc1: error: CPU you selected does not support x86-64 instruction set

13

I extracted the source code of openblas v0.3.18 Then I went into the source code directory. The following error is reported when the "make" command is executed: **cc1: error: CPU...

yejianbang

How to make sure to get exactly the same numerical results on Intel and AMD machines?

6

I am using OpenBLAS 0.3.25 version in a Docker container and wonder which build and/or runtime configuration I should use to make sure to get exactly the same results on...

VInbv

OpenMP locks instead of busy-waiting with NUM_PARALLEL

13

Brief description: This pull request proposes replacing MAX_PARALLEL_NUMBER/NUM_PARALLEL feature with OpenMP locking mechanism to achieve a more refined implementation in line with OpenMP specifications and consistent with Pthreads and Win32...

shivammonaka

Parallelized dgetrf doesn't detect matrix as singular

9

I have the following matrix: 10,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,10,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,10,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,5.6353902507748659,0,0,-5.6353902507748659,0,0,0,0,0,0,0, 0,0,0,0,5.6353902507748659,0,0,-5.6353902507748659,0,0,0,0,0,0, 0,0,0,0,0,5.6353902507748659,0,0,-5.6353902507748659,0,0,0,0,0, 0,0,0,-5.6353902507748659,0,0,15.635391250774864,0,0,-10,-9.9999999999999995e-07,0,0,0, 0,0,0,0,-5.6353902507748659,0,0,25.635391250774866,0,-10,-10,-9.9999999999999995e-07,0,0, 0,0,0,0,0,-5.6353902507748659,0,0,25.635390250774865,-10,0,-10,0,0, 0,0,0,0,0,0,-10,-10,-10,30,0,0,0,0, 0,0,0,0,0,0,-9.9999999999999995e-07,-10,0,0,20.000001999999999,-10,-9.9999999999999995e-07,0, 0,0,0,0,0,0,0,-9.9999999999999995e-07,-10,0,-10,30.001000860019595,0,-10.000999860019597, 0,0,0,0,0,0,0,0,0,0,-9.9999999999999995e-07,0,5.635391250774866,-5.6353902507748659, 0,0,0,0,0,0,0,0,0,0,0,-10.000999860019597,-5.6353902507748659,15.636390110794462, ...which should be singular. At least getting its inverse in GNU...

jmtilli

If application has sgemm without underscore how do we handle in Openblas

5

If application has sgemm without underscore how do we handle in Openblas? Also is it possible to choose which version of the functions are created - AKA sgemm vs sgemm_?...

ChipKerchner

Aliasing in `geadd`

1

Fortran has a function logic in which arrays are assumed not to alias with each other, but many BLAS functions like `ddot` nevertheless still work correctly with aliasing in the...

david-cortes

Compilation Error on Sapphire Rapids: Possible Parallel Build Conflict

3

I'm encountering a compilation issue with OpenBLAS on the Sapphire Rapids platform, using the latest code from the develop branch. My compilation command is: ``` make CC=gcc FC=gfortran BINARY=64 USE_OPENMP=1...

sandlbn

sched_yield spam

16

Openblas spams (40k interrupts/s) the sched_yield syscall, resulting in severe performance degradation. Highest impact: > /usr/lib/x86_64-linux-gnu/libc-2.31.so(__sched_yield+0x7) [0xe14e7] > [...]/openblas/linux-x86_64/libopenblas_nolapack.so.0(sgemm_thread_nt+0xe35) [0x1f4c75] > [...]/openblas/linux-x86_64/libopenblas_nolapack.so.0(exec_blas+0xce) [0x348e8e] > [...]/openblas/linux-x86_64/libopenblas_nolapack.so.0(sgemm_thread_nt+0x4f3) [0x1f4333] > [...]/openblas/linux-x86_64/libopenblas_nolapack.so.0(sgemm_thread_tn+0x117) [0x1f5097] >...

Jpx3

Test kernel_regress:skx_avx fails on RISC-V platform

7

Environment: OpenBLAS version: release 0.3.26 OS: revyos CPU: Sophgo sg2042， RISC-V rv64imafdc with rvv 0.71 Compiler: g++ 10.4, THead version. https://github.com/revyos/gcc/tree/revyos-gcc10.4-thead-dev Compile command: `make HOSTCC=gcc-10 TARGET=C910V CC=riscv64-linux-gnu-gcc-10 FC=riscv64-linux-gnu-gfortran-10 -j 64`...

leavelet