llama.cpp release b10017: Increased buffer size for SYCL USM allocations.

Core summary

One sentence to understand this update

The llama.cpp project released version b10017, increasing the minimum buffer size for SYCL USM system allocations from 1 GiB to 4 GiB to improve memory handling for large models.

Impact & opportunity

What this could mean

This update could enhance the stability and performance of llama.cpp when running larger models on SYCL-enabled hardware, benefiting developers working with local LLMs.