BLAST RADIUS

Open-weight & owned stack · Quantization efficiency

llama.cpp Improves SYCL Multi-GPU Performance

Nov 2, 2026, first seen via Training & inference tooling releases

llama.cpp updates added SYCL communication optimizations for Intel Arc Pro B70 GPUs and reported 19% higher prompt throughput.

Read the original at github.comOpens the publisher's site in a new tab

More in Open-weight & owned stack

Meituan Releases LongCat-2.5 Multimodal ModelNov 2Dejan reproduces latent skills in GemmaNov 2Reflection AI Releases Beam Open-Weight ModelNov 2Google Meta Open-Source Native PyTorch TPU BackendNov 2DeepSeek adds Creator Mode for custom pluginsNov 2