Tags: Pqlet/ao

use torch.acceraltor.current_accelerator()

Add test/test_low_bit_optim.py to XPU CI

enable inference/bench_float8_inference.py benchmark for xpu

[xpu][mx][test] Enable multicard tests for xpu

Signed-off-by: Ula Golowicz <urszula.golowicz@intel.com>

enable microbenchmarks for xpu

[mx][test] Enable non-cuda in test_mx_serialization

Signed-off-by: Ula Golowicz <urszula.golowicz@intel.com>

[mx][test] Enable non-cuda in test_mx_mm

- Replace hardcoded device="cuda" with torch.accelerator
- Use F.scaled_mm (torch 2.10+) which takes explicit
  ScalingType/SwizzleType args enabling non-CUDA backends,
  fall back to torch._scaled_mm on older versions
  (CUDA-only, implicit layout).
- Skip xpu tests when oneDNN < 3.11

Signed-off-by: Ula Golowicz <urszula.golowicz@intel.com>

[mx][test] Enable non-cuda in test_mx_tensor

- Replace hardcoded device="cuda" with torch.accelerator
- Add prameter ids for readable test names.

Signed-off-by: Ula Golowicz <urszula.golowicz@intel.com>

[mx][test] Enable non-cuda in mx_formats/test_kernels

- Replace hardcoded device="cuda" with torch.accelerator
- Add prameter ids for readable test names.

enable float8/float8_roofline.py benchmark for xpu