llm theoretical performance analysis tools and support params, flops, memory and latency analysis.
-
Updated
Jul 11, 2025 - Python
#
gpu-performance
Here are 40 public repositories matching this topic...
code for benchmarking GPU performance based on cublasSgemm and cublasHgemm
-
Updated
May 20, 2022 - Cuda
Hands-on Machine Learning Infrastructure on Kubernetes. Using Microk8s/Ubuntu on Paperspace Cloud.
-
Updated
Mar 2, 2024 - Python
A systematic CPU/GPU performance study of lightgbm and xgboost classifiers for different data shapes and hardware setups.
-
Updated
Aug 12, 2020 - Jupyter Notebook
Windows toolkit for improving NVIDIA RTX GPU performance across generations, with profile-based UEFI configuration, hardware inventory, and CUDA/cuBLAS benchmarks.
windows benchmarking gpu cuda windows-10 nvidia uefi efi rtx nvidia-cuda nvidia-gpu gpu-performance windows-11 hardware-research gpu-performance-engineering
-
Updated
Nov 2, 2026 - C
Disable GPU Thermal and Change GPU Governor to performance. (Only for snapdragon devices).
-
Updated
Nov 25, 2023 - Shell
gpu thrashingNVIDIA GPU Unified Memory diagnostic tool — architecture-aware, measurement-based, PCIe/coherent transport detection
linux cuda memory-analysis oom nvidia-gpu page-faults gpu-performance cupti dgx unified-memory gpu-vram gpu-diagnostics dgx-spark cuda-performance gpu-thrashing cuda-malloc-managed register-pressure memory-migration fault-burst-ratio uma-diagnostics
-
Updated
May 4, 2026 - Cuda
Evidence-driven PresentMon diagnostics and policy modeling paired with bounded owned-lab D3D11 runtime actuation.
windows open-source runtime d3d11 performance-analysis vram gpu-performance memory-optimization presentmon cpu-gpu frame-pacing readback frame-pipeline gpu-upload updatesubresource
-
Updated
Oct 30, 2026 - Python
Agent Skills and an MCP server for GPU performance profiling, benchmarking, optimization, and reporting, with an inference focus.
kubernetes benchmarking performance gpu mcp cuda slurm inference codex gpu-performance llm vllm llm-inference gpu-profiling mcp-server agent-skills claude-code cuda-profiling
-
Updated
Nov 9, 2026 - Python
This repository provides the latest benchmarks for the CHARMM/pyCHARMM program on GPUs
-
Updated
Jan 6, 2025 - Rich Text Format
Interactive theoretical Kimi-K3 inference roofline calculator for H200, B300, and GB300
-
Updated
Aug 11, 2026 - Python
Comprehensive performance analysis of DeepSeek V3 quantization levels (FP16, Q8_0, Q4_0) on 16GB GPU environments.
quantization model-evaluation fp16 gpu-performance latency-analysis model-quantization inference-acceleration model-optimization llm-inference llm-optimization deepseek-v3 throughput-analysis
-
Updated
Sep 27, 2025 - Python
Windows runtime for read-only process monitoring, cooperative D3D11/D3D12/Vulkan execution and bounded shared-memory research. C++20 + .NET, explicit safety gates and reproducible evidence.
windows open-source performance ram runtime dotnet vulkan d3d12 d3d11 shared-memory systems-programming content-cache vram cpp20 gpu-performance memory-optimization readback gpu-upload updatesubresource
-
Updated
Nov 6, 2026 - C#
📊 Mobile VR Performance Optimization Project
udacity csharp unity optimization unity3d virtual-reality android-app google-cardboard-vr cpu-profiling msaa udacity-nanodegree mobile-development google-vr-sdk mobile-vr gpu-performance
-
Updated
Feb 20, 2018 - C#
Cycle-accurate UMA fault latency and bandwidth measurement for NVIDIA GPUs. C and PTX. No Python. Pascal (SM 6.0) through Blackwell GB10 (SM 12.1).
pascal cuda nvidia bandwidth aarch64 rtx uma ptx gpu-performance memory-profiling cuda-c unified-memory gb10 gpu-diagnostics dgx-spark rrace-blackwell gpu-proformance fault-latency
-
Updated
Apr 28, 2026 - Cuda
Reproducible long-context inference benchmark comparing vLLM, SGLang, and TensorRT-LLM on NVIDIA GB10.
benchmarking cuda nvidia gpu-performance long-context vllm llm-inference tensorrt-llm sglang prefix-caching
-
Updated
Sep 12, 2026 - Python
Empirical LLM inference performance benchmarks: vLLM vs llama.cpp vs NVIDIA Triton across A100 & T4 GPUs
benchmarking performance gpu triton benchmarks throughput gpu-performance triton-inference-server llm llamacpp llama-cpp vllm
-
Updated
Oct 18, 2026 - Python
Single-head CUDA attention kernel: naive SDPA --> fused softmax --> occupancy-tuned variants, benchmarked against cuDNN SDPA with Nsight Compute profiling.
-
Updated
Apr 20, 2026 - Cuda
Python lab for exploring memory bandwidth, cache effects, and locality in accelerator workloads
performance-engineering hpc cuda memory-bandwidth gpu-performance roofline-model cache-locality tiling-optimization
-
Updated
Dec 24, 2025 - Python
Sparse binary 2D FFT on CUDA/cuFFT with memory-footprint optimization, streaming tiles, Hermitian symmetry, and Nsight analysis.
-
Updated
May 2, 2026 - Cuda
Add this topic to your repo
To associate your repository with the gpu-performance topic, visit your repo's landing page and select "manage topics."