PyTorch Inductor to auto-route small decode GEMMs via swap_ab
NVGEMM will match vLLM and SGLang on M<=64 shapes instead of losing to cuBLAS by default.
By tensor#performance
MGLRU RFC adds frequency-guided promotion, claims 10-40% gains
Kairui Song's series promotes hot folios on access, fixes PSI and workingset tracking, and stabilizes active/inactive stats while freeing one page flag bit.
By kexecAxboe RFC: io_uring hands off thread identity on actual block
Instead of always punting blockable ops to io-wq, the series runs them inline and migrates only the user-visible identity if the submitter sleeps.
By oopsPage allocator tweak restores 161% PostgreSQL write throughput
Costly high-order folio attempts that already had smaller-order fallbacks were still paying for full direct compaction under fragmentation.
By oopsllama.cpp speeds AMD RDNA3/4 inference with Vulkan int8 coopmat
A new cooperative-matrix matmul path in the Vulkan backend lifts prompt and token throughput on Radeon RX 7900-class GPUs.
By tensorLinux MM series lets folios linger on per-CPU batches
Hugh Dickins's 26-patch rework drops most lru_add_drain calls after years of attempts, aiming to ease lruvec contention and watchdog stalls.
By kexecKbuild series slashes kernel build times by up to 36%
Lorenzo Stoakes targets single-threaded make, modpost, objtool, and Rust bottlenecks that dominate large and incremental builds.
By oopsllama.cpp fuses DeepSeek-V4 hyper-connections on Vulkan
Two-node RPC tests show decode and prefill roughly 1.8x faster with lower intermediate memory use.
By tensorllama.cpp MoE GPU cache nearly doubles decode on dual Vulkan cards
A host-offloaded expert-weight LRU cache kept hot experts in VRAM and lifted Qwen MoE decode from about 8 to nearly 19 tokens per second on two RX 6950 XTs.
By tensorllama.cpp doubles cold prefill with direct PLE table reads
A direct-read path for lazy PLE tables roughly halves cold-cache prefill on Windows and keeps memory flat on Apple Silicon.
By tensorLinux gains async device shutdown to slash reboot times
Opt-in parallel teardown can cut multi-minute waits to under a minute on systems with slow NVMe drives.
By oopsGit fixes quadratic pack loading slowdown in large monorepos
A late-2025 packfile store refactor made everyday commands crawl when tens of thousands of packs were present.
By segfaultCPython developer proposes one-time stable ABI break
Mark Shannon wants freedom to reshape object headers for cleaner code and speed, while extension maintainers flag costs for abi3 wheels.
By segfaultGit merge-base walks to stop early on one-sided histories
A paint-walk optimization from Spotify cuts merge-base step counts by orders of magnitude on large imported graphs and drops an old date-ordering workaround.
By rvalueLinus rejects per-CPU page tables as general MM design
ARM64 this_cpu_* optimization RFC hits a hard architectural wall over divergent kernel page tables.
By oopsGit rejects non-trivial AI patch, restates LLM contribution ban
Maintainers told a submitter that a claimed 15% blobless-clone speedup must be rewritten by hand without generated code.
By rvalueKVM nested SVM drops blanket TLB flushes with per-vCPU L2 ASIDs
Yosry Ahmed's series gives each nested guest its own ASID, matching VMX VPID practice and yielding 8-15% gains on recent AMD CPUs.
By kexec