Tartan: Evaluating Modern GPU Interconnect via a Multi-GPU Benchmark Suite

NUMA-aware multi-CPU multi-GPU data transfer benchmarks

競馬データを閲覧できるソフト「KMY競馬」を開発します

This script collects some informations about NVLink and PCI bus traffic of NVidia GPUs. Results are published as prometheus metrics via a websocket.

Ares: Multi-Cluster Kubernetes Scheduler with GPU Topology Optimization (Intra-Node, Inter-Node, Inter-Cluster) and Exactly-Once Execution Semantics

GPU-native agent-swarm orchestration for the NVIDIA AI stack — NeMo, NIM, Triton, DCGM, NGC, NIXL, OpenShell. Spawn GPU-pinned agent teams across DGX/HGX nodes with NVLink-aware scheduling, task DAGs, adaptive scheduling, and full observability.

Close-to-metal C/CUDA lab for RL inference fast paths: persistent GPU workers, hugepage KV arenas, cacheline-aware command rings, and async reward handoff. Goal: remove page faults, malloc/free, scheduler wakeups, CPU round-trips, and KV migration from the per-token path.

C++ command-line tool for managing NVIDIA Fabric Manager partitions. Supports non-interactive mode and advanced partition operations.

Open hardware desktop AI node: 4× Tesla V100, 128GB HBM2, PCIe/NVLink topology and V-Core liquid/air cooling.

V-AXION-512: Dual-Tier Post-Entropic Framework. I. PROTOCOL: SR-512, GS-512 & G-STORM-512 for O(1) deterministic state recovery. II. ECOSYSTEM: NEPTUNE-PHX, PHX-BUSLINK, KALMAN-ANCHOR, PHX-GENESIS, DIRECT-FABRIC & AETERNA-FLUX for Sigma-H energy harvesting. — Juho Artturi Hemminki (2026)

What to consider when running AI Inference at scale on Kubernetes

CORTEX - Hardware-Aware AI Workload Scheduler

A hybrid testbed for evaluating top open-source LLMs (like gpt-oss-20b and Llama 3.3) on local, cloud GPUs, and AWS Inferentia2/Trainium instances, focusing on vLLM optimization, capacity management, kernel bypass, hardware-software co-design, as well as supporting infrastructure such as NCCL, RDMA, NVMeoF.

Comprehensive NCA-AIIO exam prep: study notes, diagrams, screenshots, and field experience for the NVIDIA Certified Associate: AI Infrastructure and Operations certification.

NCCL collective benchmarks on an 8×H100 NVSwitch host — busbw vs link budget, NVLS/Ring/Tree, small-message latency floors (eager vs CUDA Graph vs symmetric memory), and the TP-decode comms ceiling they imply. Includes a quiet-box rerun methodology for attribution.

Open-source stencil-aware multi-GPU Conjugate Gradient solver on 8× A100 NVLink. 2.07× SpMV vs cuSPARSE · 1.44× above NVIDIA AmgX · 93.5% strong scaling efficiency. Profiled with Nsight Systems & Nsight Compute.

Provide open-source access to detailed AI hardware specs, benchmarks, and infrastructure data for informed decision-making and analysis.