-
I was trying to use HIP plugin (via `ONEAPI_DEVICE_SELECTOR=ext_oneapi_hip:gpu`) and got an error from UR plugin though. Was the SYCL for HIP now defaulted to use UR plugin or if there is a way to swi…
-
When I compile with `--hipsycl-targets=hip:gfx900` I am getting warning:
```
[hipSYCL Warning] backend_loader: Could not load backend plugin: /opt/hipSYCL/bin/../lib/hipSYCL/librt-backend-cuda.so
…
-
Hi,
I've just built adaptivecpp on Nvidia GPU and then built PortBLAS. Compared the benchmarks to dpcpp.
```bash
git clone --recursive https://github.com/codeplaysoftware/portBLAS.git
cd portB…
-
Hi,
I download the daily built binary and then set some env like
```bash
export PATH=$DPCPP_HOME/bin:$PATH
export LD_LIBRARY_PATH=$DPCPP_HOME/lib:$LD_LIBRARY_PATH
```
but run into several compil…
-
**Describe the bug**
In its default "direct mode", `ccache` examines the preprocessor output to determine all dependencies for a given translation unit. In doing so, `ccache` gets tripped up by the…
psalz updated
1 month ago
-
When compiling the following minimal reproducer:
```
#include
using namespace sycl;
int main() {
queue q{};
q.submit([&](handler &cgh) {
cgh.parallel_for(
sycl::nd_ran…
-
Hi,
I was trying to run some code on an Nvidia GPU. However, I've noticed that the performance is greatly reduced in comparison to the CUDA native counterpart.
At this point, I used the nvprof and…
-
### Describe the bug
When compiling AOT for a specific target the corresponding macro is set to 1, while the macros for all the other targets are set to 0. However, for the CUDA backend, the macro th…
-
**System information**
- OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Linux Ubuntu 16.04
- TensorFlow installed from (source or binary): source
- TensorFlow version: git master
- Pytho…
-
Hi All,
first of amazing work !!!!
I am really impressed and happy that there is a open source project out there pushing SYCL.
I am at the moment trying to implement this as our goto solution …