-
## Description
Replace all HIP calls with `hipper` calls. Continues #427.
## Motivation and context
HIP is complex, out of our control, gets in the way, and often breaks things. There is …
-
I'm currently trying to assert the status of expectation we can have on this setup for DeepSpeech, relying on TensorFlow with ComputeCpp. I have been able to cross-build the driver, and most of the `T…
-
Imagine a device-side function with the following signature:
```c++
void foo(int* vec);
```
I don't know if `vec` comes from global, local, constant or private memory. However, inside `foo` I'd …
-
C++ modules do not seem to currently work with intel/llvm when SYCL is used. I'm curious if this is likely to be added in the near future, or if there are architectural reasons that prevent C++ modul…
-
### Describe the bug
The initial value for `sycl::minimum` is set to `inf` which gets turned into `0` with `-ffast-math`. More generally, both min and max are affected for any type which `has_infinit…
-
Some math manglings from xocc's SPIR libraries appear to be incorrect or missing. In the sense that math functions correctly translated to their SPIR mangled names (found in: https://github.com/Khrono…
-
This Issue is being raised as an RFC to contribute SYCL backend to the current clojureCUDA backend of BAYADERA.
This SYCL backend was developed using Intel's [oneAPI ](https://www.intel.com/content…
-
Hi!,
I have the following question: _generic_ is excellent for compiling once and targeting multiple devices, but I am noticing that the performance is lower compared to when I compile specifying …
-
```
C:\Users\akonoval\src\include_hext>clang++.exe -v -I include\oneapi\dpl\stdpar -I include -O0 -g -w -S test.cpp
clang version 17.0.0 (https://github.com/intel/llvm 1a5e53216b009177c73dfd6d644a5c…
-
### 🐛 Describe the bug
```
import torch
assert torch.xpu.is_available(), "Intel XPU is not available"
batch_size = 4
vocab_size = 4
# RuntimeError: Required aspect fp64 is not suppor…