Name and Version
root@dcd9fcec754d:/app# ./llama-cli --version
load_backend: loaded SYCL backend from /app/libggml-sycl.so
load_backend: loaded CPU backend from /app/libggml-cpu-haswell.so
version: 8248 (5f4cdac38)
built with IntelLLVM 2025.2.1 for Linux x86_64
Operating systems
Linux
Which llama.cpp modules do you know to be affected?
No response
Command line
Problem description & steps to reproduce
llama.cpp SYCL is using the CPU 99% of the time for BF16 models. F16 models work fine on the GPU
First Bad Commit
No response
Relevant log output
Name and Version
Operating systems
Linux
Which llama.cpp modules do you know to be affected?
No response
Command line
Problem description & steps to reproduce
llama.cpp SYCL is using the CPU 99% of the time for BF16 models. F16 models work fine on the GPU
First Bad Commit
No response
Relevant log output