-
With oneDNN and ACL tightly coupled for aarch64 builds, having them in two repos is making it harder to keep the versions in sync. The request here is to bring ACL sources into oneDNN, may be as a git…
-
Hi, thanks for this benchmark.
Are you planning on including **ARM Compute Library** in future version?
-
It seems that `configure` needs to be called per each `run` for `NEGEMM`.
Because of this issue we have work around in oneDNN fork: https://github.com/openvinotoolkit/oneDNN/commit/a5ca47f7860b3cc44b…
-
NEDeconvolutionLayer `run()` with f16 tensors takes more time than NEDeconvolutionLayer `run()` with f32 tensors.
On Ampere f32 version takes ~66 milliseconds, f16 version ~80 milliseconds.
ACL bu…
-
Build fail by build toolchain version
```
In file included from ./arm_compute/core/TensorShape.h:29,
from ./arm_compute/core/PyramidInfo.h:27,
from src/core/Pyr…
-
Benchmark details: 1d convolution of a 2^16 wide 1D input signal with a length 3 kernel. Both input and output channels are 1. There is no bias term.
```
% strings arm_compute-v24.06-bin-android-a…
-
With GCC 15, the build fails due to missing cstdint include. GCC 15 removed some extra includes in the standard headers and caused the issue. See https://gcc.gnu.org/gcc-15/porting_to.html for more in…
-
### What Happened
When I attempt to build TVM-Unity I run into errors with the ACL Lib. I believe/hope it is because of this change not being reflected within the TVM ACL CMake Config.
The 2 er…
-
I am having problem with enabling quanteda's parallel computing in R on M3 Macbook. In GitHub, the readme says:
>Windows or macOS users do not have to install TBB or any other packages to enable pa…
-
**Output of 'strings libarm_compute.so | grep arm_compute_version':**
arm_compute_version=v24.02.1 Build options: {'neon': '1', 'opencl': '0', 'openmp': '0', 'cppthreads': '1', 'arch': 'armv8.6-a', '…