Highly-available Distributed Fault-tolerant Runtime
-
Updated
Apr 22, 2023 - Rust
Highly-available Distributed Fault-tolerant Runtime
A community-oriented list of useful NUMA-related libraries, tools, and other resources
High performance and low latency Exchange Collections written in Java
Local-affinity first NUMA-aware allocator with optional fallback.
Pterodactyl Docker Images
A low-latency Java 21 graph runtime for bounded in-process pipelines, explicit backpressure, stage fusion, and NUMA-aware topology diagnostics.
Experimental NUMA-aware, lock-free heap memory allocator
Topology-aware Kubernetes scheduler for multi-tenant, heterogeneous clusters
A blazing-fast, NUMA-aware memory allocator written in pure Rust.
gonuma: A utility library for writing NUMA-aware Go applications - please use https://github.com/lrita/numa instead.
[EXPERIMENTAL] Custom NUMA node scheduler - distributes non-adapted Windows programs into several NUMA nodes evenly
This project implements two types of thread pools that run on macOS and Linux. The NormalThreadPool is compatible with both platforms, while the NumaThreadPool (NUMA-aware) is only available on Linux due to macOS's lack of NUMA support (Apple Silicon applies UMA arch).
Memory-bound Jacobi stencil on SuperMUC-NG, optimized from AVX-512 and temporal blocking to NUMA-aware OpenMP and hybrid MPI. Explained by LIKWID hardware counters.
NUMA-first runtime for latency-critical Rust applications.
Deploy hardened, isolated workerd environments for secure multi-tenant self-hosting.
Ultra-fast JSON minifier. Multiple performance modes, and cross-platform language bindings.
🖥️ Streamline your workflow with bast, the AI-powered terminal assistant that converts plain English tasks to shell commands instantly.
M.Eng. thesis authored during my studies @cse.uoi.gr
CPU-first inference orchestration for multi-socket x86. Validates NUMA-aware memory control and heuristic routing can deliver ≥25% throughput uplift on large models without retraining.
XR-optimized Linux kernel source tree for Bigscreen Beyond 2e on AMD GPUs on Numa aware, multi-processor BCI servers ⚡️
To associate your repository with the numa-aware topic, visit your repo's landing page and select "manage topics."