Tag
NVIDIA introduces CUDA Rust with two tracks (SIMT and Tile) for writing GPU kernels natively in Rust, enabling performance and developer experience improvements in AI systems.
Linus Torvalds shares his thoughts on the use of large language models in Linux kernel development, likely expressing caution regarding their impact on code quality and community processes.
Linux 7.2 has finally removed the deprecated strncpy API after six years and over 360 patches due to its bug-prone nature, replacing it with safer alternatives like strscpy and memcpy.
Google DeepMind's pre-training lead Vlad Feinberg highlights kernel development and low-level performance engineering as high-demand skills for frontier AI labs.
Hugging Face is inviting kernel developers to publish their custom kernels on the Hugging Face Hub via 🤗 Kernels, offering benefits like consistent build structure, ease of use, standardized distribution, and reproducibility.
This article covers recent improvements and future plans for the Linux kernel's swap subsystem, including reduced per-page overhead, folio-based helpers, and efforts to make swapping more friendly to solid-state storage, as discussed at the 2026 Linux Storage, Filesystem, Memory Management, and BPF Summit.
A technical overview of implementing an NVMe driver for the Maestro operating system, covering PCIe interface, queues, and memory-mapped I/O details.
Paul McKenney discusses unconventional and corner-case RCU (Read-Copy-Update) implementations, including timed-wait RCU approaches used in early Unix systems and fixed-buffer RCU concepts related to memory quarantining, illustrating creative but potentially dangerous synchronization techniques in kernel development.