Commits

Commits on Mar 24, 2025

Update CMakeLists.txt
Show description for 4955049
AlexandreZhdanov
authored
4955049
Copy full SHA for 4955049

Commits on Mar 21, 2025

Add Lattice Boltzmann Fluid Simulation Example (#3455)
Show description for c6269a6
edwinsolisf
authored
c6269a6
Copy full SHA for c6269a6

Commits on Mar 14, 2025

updated 3.9 install and getting started instructions (#3496)
Show description for 360fefb
melonakos
authored
360fefb
Copy full SHA for 360fefb
Fixes for compiling with oneAPI 2025 on Linux (#3643)
Show description for b568f87
christophe-murphy
authored
b568f87
Copy full SHA for b568f87

Commits on Mar 11, 2025

bump version to 3.10

verstatx
authored and
syurkevi
committed
7b82364
Copy full SHA for 7b82364
add missing s8 tests

verstatx
authored and
syurkevi
committed
b0be72d
Copy full SHA for b0be72d
add s8 to documentation

verstatx
authored and
syurkevi
committed
0e259cc
Copy full SHA for 0e259cc
define s8 interface for AF_API_VERSION 310
Show description for 42dfafa

verstatx
authored and
syurkevi
committed
42dfafa
Copy full SHA for 42dfafa
skip rotate_linear tests for s8
Show description for 3eadfdc

verstatx
authored and
syurkevi
committed
3eadfdc
Copy full SHA for 3eadfdc
skip Richardson-Lucy test for s8
Show description for 1a0c305

verstatx
authored and
syurkevi
committed
1a0c305
Copy full SHA for 1a0c305
fix image loading for s8 tests

verstatx
authored and
syurkevi
committed
1636b5a
Copy full SHA for 1636b5a
signed 8-bit integer support

verstatx
authored and
syurkevi
committed
cdbbc75
Copy full SHA for cdbbc75

Commits on Mar 7, 2025

Use offsets for CSR/COO to dense conversion in OpenCL and oneAPI (#3633)
Show description for e073df6
christophe-murphy
authored
e073df6
Copy full SHA for e073df6
Add f16 support for modulus and norm (#3258)
Show description for 408b504

authored
408b504
Copy full SHA for 408b504

Commits on Feb 26, 2025

Join does not always respect the order of provided parameters (#3511) (#3513)
Show description for 48b7a9e

willyborn
and
christophe-murphy
authored
48b7a9e
Copy full SHA for 48b7a9e

Commits on Feb 21, 2025

Add support for CUDA 12.7 and 12.8 (#3636)
FloopCZ
authored
d63c391
Copy full SHA for d63c391

Commits on Feb 20, 2025

Fix race condition in OpenCL kernel (#3535)
Show description for 6cea4d3
frasercrmck
authored
6cea4d3
Copy full SHA for 6cea4d3
The test cmake file needs to be updated with the commit hash which includes the new test data needed for the test added in pull requests #3585 and #3587. (#3635)
christophe-murphy
authored
ccabfe6
Copy full SHA for ccabfe6

Commits on Feb 17, 2025

fix fallthrough in reduce_by_key_common for u8 (#3503)
Show description for 553c38d
verstatx
authored
553c38d
Copy full SHA for 553c38d

Commits on Feb 13, 2025

3539 build oneapi version 2024 incompatible for mkl (#3573)
Show description for b1e85d3

christophe-murphy
and
edwinsolisf
authored
b1e85d3
Copy full SHA for b1e85d3

Commits on Feb 11, 2025

Fix issue 3525: correct handling of array indexing with sequence (#3585)
Show description for 18028c0
edwinsolisf
authored
18028c0
Copy full SHA for 18028c0
Added testing for sequence indexing with non-unitary steps mixed in with array indexing (#3587)
edwinsolisf
authored
c644e4f
Copy full SHA for c644e4f

Commits on Jan 30, 2025

3560 bug incorrect results when using the pow function with float16 arguments with cuda backend (#3561)
Show description for ffda1b6
christophe-murphy
authored
ffda1b6
Copy full SHA for ffda1b6

Commits on Jan 16, 2025

Reverted an error in the interop_cuda example code where the cuda stream id was being used instead of the arrayfire stream id. (#3594)
christophe-murphy
authored
6e5dca4
Copy full SHA for 6e5dca4
Correct the conversion from float/double to half on CUDA (#3627)
willyborn
authored
eef5773
Copy full SHA for eef5773

Commits on Jan 14, 2025

Extend test for convolve2GradientNN function to verify zero padding fix in PR 3519 (#3631)
christophe-murphy
authored
5c2ea29
Copy full SHA for 5c2ea29

Commits on Jan 13, 2025

Fixed padding comparison in convolve2GradientNN (#3519)
Show description for 7127a0b

errata-c
and
errata-c
authored
7127a0b
Copy full SHA for 7127a0b
Alternative OpenCL kernel for performing the CSC matrix vector multiply using atomic operations. Benchmarking so far has shown it to be on par with the CUDA backend on my Nvidia RTX 4060 GPU. Note …
Show description for e770c88
christophe-murphy
authored
e770c88
Copy full SHA for e770c88

Commits on Jan 9, 2025

3580 bug investigate test failures when running with cuda 126 (#3588)
Show description for f4edcf2

christophe-murphy
and
FloopCZ
authored
f4edcf2
Copy full SHA for f4edcf2

Commits on Jan 8, 2025

3545 bug fp16 types not allowed for atan2 method (#3559)
Show description for f6559a5
christophe-murphy
authored
f6559a5
Copy full SHA for f6559a5
Added CMakeUserPresets.json to .gitignore, allow for local cmake configuration (#3520)
Show description for 424f1d6

errata-c
and
errata-c
authored
424f1d6
Copy full SHA for 424f1d6
The shfl_instrinsics header file contains wrapper routines for the warp primitives and calls the new primitives for CUDA versions greater than 9 and the old ones for older CUDA versions. The new pr…
Show description for 23a990a
christophe-murphy
authored
23a990a
Copy full SHA for 23a990a
Add condition so that cuda_* tests (including cuda_unified) are only built if AF_BUILD_CUDA is true (#3598)
christophe-murphy
authored
b25ff74
Copy full SHA for b25ff74
Add note to write() array class method documentation about copy on write behavior. (#3613)
christophe-murphy
authored
374bf97
Copy full SHA for 374bf97
1918 sparse matrix not updated in for loop in opencl (#3602)
Show description for 279d0ea
christophe-murphy
authored
279d0ea
Copy full SHA for 279d0ea