Keichi Takahashi

julia-on-tpu. This work documents a systematic investigation of the Julia programming language on Google Tensor Processing Unit (TPU) hardware, using Reactant.jl and the XLA compiler as the execution backend.

GaussianSplatting.jl. Gaussian Splatting in pure Julia

StructArrays.jl. Efficient implementation of struct arrays in Julia

slurm-bridge. Run Slurm as a Kubernetes scheduler. A Slinky project.

skills. Agent Skills for Google products and technologies

HeCBench. C++

PyNLO. Python package for nonlinear optics

benchpark. An open collaborative repository for reproducible specifications of HPC benchmarks and cross site benchmarking environments

AI-for-Science-Supercomputer. Landing page for the AI for Science Supercomputer at RIKEN R-CCS

skills. Skills for Real Engineers. Straight from my .claude directory.

skills. Public repository for Agent Skills

ai-performance-engineering. Code, labs, and resources for O'Reilly AI Systems Performance Engineering: GPU optimization, distributed training, inference scaling, and full-stack tuning.

markitdown. Python tool for converting files and office documents to Markdown.

docling. Get your documents ready for gen AI

optuna-integration. Extended functionalities for Optuna in combination with third-party libraries.

pi. AI agent toolkit: unified LLM API, agent loop, TUI, coding agent CLI

ra-skills. リサーチアドミニストレーション(RA)のスキルハブ — Claude Code skills フォーマットでRA業務知見を構造化・共有

magika. Fast and accurate AI powered file content types detection

BitNet. Official inference framework for 1-bit LLMs

ZeroKVM. A low-cost, easy-to-build KVM-over-IP device that acts as a DisplayLink monitor

mace-osaka26. Shell

cscs-reframe-tests. The CSCS ReFrame test suite

Chimbuko. Performance Analysis Framework for Scientific Workflows for online Data Analysis and Performance Visualization

JaxPM. JAX-powered Cosmological Particle-Mesh N-body Solver

integrated-matrix-extension. Administrative repository for the Integrated Matrix Extension Task Group

catalyst. A JIT compiler for hybrid quantum programs in PennyLane

mlir-tutorial. MLIR For Beginners tutorial

Stim. A fast stabilizer circuit library.

LightRAG. [EMNLP2025] "LightRAG: Simple and Fast Retrieval-Augmented Generation"

tieredsort. A fast, header-only C++17 sorting library for numeric types. 3.6x faster than std::sort, up to 21x on dense data

box64. Box64 - Linux Userspace x86_64 Emulator with a twist, targeted at ARM64, RV64 and LoongArch Linux devices

FEX. A fast usermode x86 and x86-64 emulator for Arm64 Linux

bonsai-dt. Programmable Decision Tree Framework

spectralDNS. Spectral Navier Stokes (and similar) solvers in Python

hit3d. Fortran

shenfun. High performance computational platform in Python for the spectral Galerkin method

Incompact3d. Current CPU version of our solver for the Navier-Stokes equations

pencil-code. A high-order finite-difference code for compressible hydrodynamic flows with magnetic fields and particles

AFiD. A highly parallel application for Rayleigh-Benard and Taylor-Couette flows

vpic. Vector Particle-In-Cell (VPIC) Project

libpressio_tutorial. A Tutorial for LibPressio

tiny8. A tiny CPU simulator written in Python

tiny-gpu. 🐉 A lightweight Python-based GPU architecture simulator that demonstrates how parallel threads, registers, memory, and instructions work on a GPU.

annotated_MIND. Python implementation of Manifold Inference from Neural Dynamics (MIND)

mdscuda. CUDA implementation of Multidimensional Scaling

vstl. Vectorized template library like STL

linux-aio. How to use the Linux AIO feature

python-libaio. Python wrapper for libaio

tt-sim. Tenstorrent architecture simulator

benchmarking-fft. choosing FFT library...

sip_mss_AnDeS. Anomaly Detection System (AnDeS)

llvm-project. llvm-project cloned from https://github.com/llvm/llvm-project and modified for VE

rv. RV: A Unified Region Vectorizer for LLVM

tutorials. C++

ginkgo. Numerical linear algebra software package

ADIOS2.jl. A Julia interface to ADIOS2

Cthulhu.jl. The slow descent into madness

julia-hpc-tutorial-icpp25. Jupyter Notebook

Triton-Puzzles. Puzzles for learning Triton

ssd-gpu-dma. Build userspace NVMe drivers and storage applications with CUDA support

donard. Donard: A PCIe Peer-2-Peer kernel patch and library that builds on top of NVM. Express. Also see https://github.com/sbates130272/linux-donard for the linux kernel source used with the user-space code and modules.

PCIconf. Most popular PCI devices and pciconf reports

legate. The Foundation for All Legate Libraries

jitify. A single-header C++ library for simplifying the use of CUDA Runtime Compilation (NVRTC).

libmp. Simple message passing library

numba-cuda. The CUDA target for Numba

MIT-Supercloud-Dataset. Repository for the MIT Supercloud Dataset

CompilerGym. Reinforcement learning environments for compiler and program optimization tasks

PolyBenchC-4.2.1. PolyBench/C benchmark suite (version 4.2.1 beta) from http://web.cse.ohio-state.edu/~pouchet/software/polybench/

sapporo-service. A standard implementation conforming to the Global Alliance for Genomics and Health (GA4GH) Workflow Execution Service (WES) API specification.

TQSim. C++

arpack-ng. Collection of Fortran77 subroutines designed to solve large scale eigenvalue problems.

container. A tool for creating and running Linux containers using lightweight virtual machines on a Mac. It is written in Swift, and optimized for Apple silicon.

containerization. Containerization is a Swift package for running Linux containers on macOS.

code_contests. C++

arkouda. Arkouda (αρκούδα): Interactive Data Analytics at Supercomputing Scale :bear:

cuvs. cuVS - a library for vector search and clustering on the GPU

dopamine. Dopamine is a research framework for fast prototyping of reinforcement learning algorithms.

Trilinos. Primary repository for the Trilinos Project

hpc-workspace-v2. location for development of next major version of hpc-workspace

hpc-workspace. Automatically exported from code.google.com/p/hpc-workspace

wenyan. 文言文編程語言 A programming language for the ancient Chinese.

vfkit. Go

cvw. CORE-V Wally is a configurable RISC-V Processor associated with RISC-V System-on-Chip Design textbook. Contains a 5-stage pipeline, support for A, B, C, D, F, M and Q extensions, and optional caches, BP, FPU, VM/MMU, AHB, RAMs, and peripherals.

system-prompts-and-models-of-ai-tools. FULL Augment Code, Claude Code, Cluely, CodeBuddy, Comet, Cursor, Devin AI, Junie, Kiro, Leap.new, Lovable, Manus, NotionAI, Orchids.app, Perplexity, Poke, Qoder, Replit, Same.dev, Trae, Traycer AI, VSCode Agent, Warp.dev, Windsurf, Xcode, Z.ai Code, Dia & v0. (And other Open Sourced) System Prompts, Internal Tools & AI Models

peft. 🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.

HEBO. Bayesian optimisation & Reinforcement Learning library developed by Huawei Noah's Ark Lab

kerncraft. Loop Kernel Analysis and Performance Modeling Toolkit

kperf. A benchmarking tool to evaluate Knative performance

flamegraph. Easy flamegraphs for Rust projects and everything else, without Perl or pipes <3

caio. Linux AIO c python bindings

Shakti. Shakti will be providing developers with fast & powerful yet easy to use Python Async Interface, without the complexity of using Liburing & io_uring directly.

array-language-comparisons. A comparison of array languages & libraries: APL, J, BQN, Uiua, Q, Julia, R, NumPy, Nial, Futhark, Dex, Ivy, SaC & ArrayFire.

UnicodePlots.jl. Unicode-based scientific plotting for working in the terminal

ocs-vcp-portable. VCP Portable version for the NII Gakunin Cloud On-demand Configuration Service (OCS)

cuda-toolkit. GitHub Action to install CUDA

oras. OCI registry client - managing content like artifacts, images, packages

netbox. The premier source of truth powering network automation. Open source under Apache 2. Try NetBox Cloud free: https://netboxlabs.com/products/free-netbox-cloud/

oneapi-construction-kit. C++

prek. ⚡ A fast Git hook manager written in Rust, designed as a drop-in alternative to pre-commit, reimagined.

polars-benchmark. Python

s3-benchmark. Measure Amazon S3's performance from any location.

2024-APAC-HPC-AI. Shell

2023-APAC-HPC-AI.

2022-APAC-HPC-AI. 2022 APAC HPC-AI Competition

bloaty. Bloaty: a size profiler for binaries

physicsnemo. Open-source deep-learning framework for building, training, and fine-tuning deep learning models using state-of-the-art Physics-ML methods

elpa. A scalable eigensolver for dense, symmetric (hermitian) matrices (fork of https://gitlab.mpcdf.mpg.de/elpa/elpa.git)

ollama. Get up and running with Kimi-K2.6, GLM-5.1, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.

pytest-grpc. Allow test gRPC with pytest

smol. A small and fast async runtime for Rust

ml-depth-pro. Depth Pro: Sharp Monocular Metric Depth in Less Than a Second.

vortex. An extensible, state-of-the-art framework for columnar compression, and the fastest FOSS columnar file format. Formerly at @spiraldb, now an Incubation Stage project at LFAI&Data, part of the Linux Foundation.

LibAFL. Advanced Fuzzing Library - Slot your Fuzzer together in Rust! Scales across cores and machines. For Windows, Android, MacOS, Linux, no_std, ...

angr. A powerful and user-friendly binary analysis platform!

hod-carrier. C++

Arrival-Movie-Live-Coding. Documents from a live coding session by Christopher Wolfram related to content from the 2016 film Arrival

THAPI. A tracing infrastructure for heterogeneous computing applications.

dftracer. A multi-level dataflow tracer for capturing I/O calls from workflows.

ReMPI. ReMPI (MPI Record-and-Replay)

NLP2024-tutorial-3. NLP2024 チュートリアル３ 作って学ぶ日本語大規模言語モデル - 環境構築手順とソースコード / NLP2024 Tutorial 3: Practicing how to build a Japanese large-scale language model - Environment construction and experimental source codes

zpoline. system call hook for Linux

scalene. Scalene: a high-performance, high-precision CPU, GPU, and memory profiler for Python with AI-powered optimization proposals

memray. Memray is a memory profiler for Python

vllm. A high-throughput and memory-efficient inference and serving engine for LLMs

einsumt. Multithreaded version of numpy.einsum

nvbandwidth. A tool for bandwidth measurements on NVIDIA GPUs.

cwltool. Common Workflow Language reference implementation

agios. an I/O request scheduling library at file level

libyogrt. Your One Get Remaining Time library

xsystem35-sdl2. Multi-platform implementation of AliceSoft's System3.x game engine.

TestSNAP-stdpar. C++ Parallel STL implementation of TestSNAP

kokkos-fft. A shared-memory FFT for the Kokkos ecosystem

multi-gpu-programming-models. Examples demonstrating available options to program multiple GPUs in a single node or a cluster

new-computation-principle.github.io. 次世代計算基盤に係る調査研究：新計算原理チーム

ethercrab. A pure Rust EtherCAT MainDevice supporting std and no_std environments

gpu-benches. collection of benchmarks to measure basic GPU capabilities

ChampSim. ChampSim is an open-source trace based simulator maintained at Texas A&M University and through the support of the computer architecture community.

booksim2. BookSim 2.0

openrecall. OpenRecall is a fully open-source, privacy-first alternative to proprietary solutions like Microsoft's Windows Recall. With OpenRecall, you can easily access your digital history, enhancing your memory and productivity without compromising your privacy.

PMTiles. Pyramids of map tiles in a single file on static storage

MPAS-Model. Repository for MPAS models and shared framework releases.

pyxis. Container plugin for Slurm Workload Manager

RDM-osf.io. Facilitating Open Science

PtM-Suite. PPTファイルから、スライドを抜きだし、Note部分を音声合成して動画をつくりマイクロコンテンツシステムへ登録するシステムを開発するリポジトリ

llama3-from-scratch. llama3 implementation one matrix multiplication at a time

mlx-llm. Large Language Models (LLMs) applications and tools running on Apple Silicon in real-time with Apple MLX.

scaluq. C++

dash. Data Apps & Dashboards for Python. No JavaScript Required.

bqplot. Plotting library for IPython/Jupyter notebooks

onnxscript. ONNX Script enables developers to naturally author ONNX functions and models using a subset of Python.

bohrium. Automatic parallelization of Python/NumPy, C, and C++ codes on Linux and MacOSX

hypothesis. The property-based testing library for Python

GPTune. Python

CG_on_GPU. GPU移植のための実装例（共役勾配法によるポアソン方程式求解）

nbody. GPU移植のための実装例（直接法に基づくN体計算）

rpyc. RPyC (Remote Python Call) - A transparent and symmetric RPC library for python

awesome-3D-gaussian-splatting. Curated list of papers and resources focused on 3D Gaussian Splatting, intended to keep pace with the anticipated surge of research in the coming months.

Tsunami-HySEA. Tsunami-HySEA is a module of HySEA to simulate tsunamis generated by earthquakes

tiny-cuda-nn. Lightning fast C++/CUDA neural network framework

roc-stdpar. C++

cpp_parallelization_examples. The study & production material for https://www.youtube.com/watch?v=Pc8DfEyAxzg

colmena. Library for steering campaigns of simulations on supercomputers

comm_overlap_bench. Communication/Computation Overlap Benchmark

grip. Preview GitHub README.md files locally before committing them.

capitalens. Organizing Japan's Administrative Data

adept. Automatic-Differentiation-Enabled Plasma Transport in JAX

gdbstub. An ergonomic, featureful, and easy-to-integrate implementation of the GDB Remote Serial Protocol in Rust (with no-compromises #![no_std] support)

arrayjob-runner. A tool dividing a large task list and facilitating execution with Slurm ArrayJob

K3D-jupyter. K3D lets you create 3D plots backed by WebGL with high-level API (surfaces, isosurfaces, voxels, mesh, cloud points, vtk objects, volume renderer, colormaps, etc). The primary aim of K3D-jupyter is to be easy for use as stand alone package like matplotlib, but also to allow interoperation with existing libraries as VTK.

NVFlare. NVIDIA Federated Learning Application Runtime Environment

lonboard. Fast, interactive geospatial data visualization in Jupyter.

liblbfgs. libLBFGS: a library of Limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS)

gimli. A library for reading and writing the DWARF debugging format

vistle. Software Environment for High-Performance Simulation and Parallel Visualization

py-spy. Sampling profiler for Python programs

SmartSim. SmartSim Infrastructure Library.

dakota. The core Dakota project

mlx. MLX: An array framework for Apple silicon

sip_mss_4cast. MSS data forecast system

Modern-CPP-Programming. Modern C++ Programming Course (C++03/11/14/17/20/23/26)

ZLUDA. CUDA on non-NVIDIA GPUs

ctypeslib. Generate python ctypes classes from C headers. Requires LLVM clang

Kernels. This is a set of simple programs that can be used to explore the features of a parallel platform.

metal-cpp. Metal-cpp is a low-overhead C++ interface for Metal that helps developers add Metal functionality to graphics apps, games, and game engines that are written in C++.

neko. /ᐠ. ｡.ᐟ\ᵐᵉᵒʷˎˊ˗

WfCommons. WFCommons: A Framework for Enabling Scientific Workflow Research and Development

openmp-rr.

autoray. Abstract your array operations.

ptap2dxf. PTAP2DXF - Generate CNC-cut paper tapes from .PTAP or other binaries on a home stencil cutting machine

QFI. Python

compiler. The Boa language reference compiler.

pennylane-lightning-gpu. GPU enabled Lightning simulator for accelerated circuit simulation. See https://github.com/PennyLaneAI/pennylane-lightning for all future development of this project.

cuda-quantum. C++ and Python support for the CUDA Quantum programming model for heterogeneous quantum-classical workflows

SCAMP. The fastest way to compute matrix profiles on CPU and GPU!

spatter. Benchmark for measuring the performance of sparse and irregular memory access.

sgemv-intrinsics. Optimized matrix-vector multiply for fp32 and bf16 dense matrix for SX-Aurora Vector Engine VE1 and VE2

ve-llama2.c. Inference Llama 2 in one file of pure C

quantum_software_handson.

awesome-japanese-llm. 日本語LLMまとめ - Overview of Japanese LLMs