b8072

The llama.cpp project has released version b8072, a significant update focused on expanding hardware support. The release notably adds new GPU backends for Windows, including Vulkan, SYCL (for Intel GPUs), and HIP (for AMD GPUs). It also introduces Vulkan support for Linux. This broadens the range of consumer and data center hardware capable of running local large language models (LLMs) efficiently, moving beyond the previous heavy reliance on NVIDIA's CUDA platform.