llama.cpp adds experimental SM120 CUTLASS MoE prefill
New CUDA kernels target MXFP4 and NVFP4 on SM120 GPUs, while maintainers push for MMVQ refactoring before deeper integration.
By tensor#nvidia
Nova-core moves toward stable r000 NVIDIA GSP firmware ABI
An early-review Linux series would drop versioned GSP firmware names once the still-unreleased r000 blobs ship.
By kexecVFIO patches add CXL Type-2 accelerator passthrough for guests
A 27-patch series from NVIDIA lets VMs program virtual HDM decoders and resets while the host keeps physical memory decode.
By kexecmlx5 VFIO selftest lands after four days of Claude-assisted coding
Jason Gunthorpe submitted a 10-patch series that brings ConnectX DMA and MSI coverage to the kernel VFIO tests, cut from the usual multi-month effort.
By kexec