G

gpu

Isard VDI (Virtual Desktop Infraestructure) Project. Open Source KVM Virtual Desktops based on KVM Linux and dockers. https://isardvdi.com/

Template Numerical Library https://tnl-project.org/

GPU secp256k1 toolkit on a single card: brute-force search, Pollard kangaroo for exposed pubkeys, and BIP39 seed recovery. One CUDA binary, one verified math core.

Event Generator, written in CUDA C++. Simulates LEP and LHC Collisions.

Intelligent VRAM/RAM swapping for LLM inference - Extension of KVortex | Offloading intelligent VRAM/RAM pour l'inference

https://ayinedjimi-consultants.fr

Automated LLM Benchmarking on GPU - tokens/sec, latency percentiles, VRAM profiling, multi-format support (HuggingFace, GGUF, GPTQ)

https://ayinedjimi-consultants.fr

GPU-Accelerated Hash Cracking Engine - MD5/SHA1/SHA256/SHA512/NTLM/bcrypt with CUDA PyTorch acceleration, dictionary/bruteforce/mask/combination attacks, 50+ mutation rules. For authorized pentesting only.

https://ayinedjimi-consultants.fr

LLM quantization & benchmarking on GPU - GGUF, GPTQ, AWQ, bitsandbytes | Quantification et benchmark de modeles LLM sur GPU

https://ayinedjimi-consultants.fr

Extreme KV Cache Compression for LLM Inference — C++17/CUDA implementation of TurboQuant (arXiv 2504.19874). 7.5x compression, <2% quality loss.

https://ayinedjimi-consultants.fr

GPU-accelerated embedding server for RAG systems - CUDA, FastAPI, sentence-transformers | Serveur d'embeddings GPU ultra-rapide

https://ayinedjimi-consultants.fr

PyHole is a general relativity ray tracing code written in Python and OpenCL. It includes an easy to use interface for interactive analysis of the dynamics around arbitrary axisymmetric stationary space-times.

Scholara: A Virtual Assistant Powered by LLM, RAG, and AI Agent.

Finding the right scholarship can be a timeconsuming and frustrating process. You'll likely spend hours poring over university websites, communicating with current students, and scouring the internet for information. Unfortunately, even after all that effort, there's no guarantee you'll find the perfect scholarship. That's where our AI scholar comes in. Let us handle the legwork for you, and we'll provide personalized assistance to find the most suitable scholarship for you and answer all your questions about the scholarships that interest you. We solved this by using LLM to provide a human-like when answer the students, RAG we build our own vector database to get more suitable scholarships from it based on student information and desire, also to keep our system up-to-date with all the updated information about the scholarships we built AI Web agent this helps the system to be aware about all the updates in the scholarships information, and new scholarships.

NnmfPack was conceived for efficiently solving Non-negative matrix factorization problems. NnmfPack is a collection of numerical routines written in C/CUDA. NnmfPack was designed and optimized for shared memory architectures.

Query nvidia-smi inside python and read output as list of dictionaries.

Model collisions between cells using a phase field algorithm implemented using CUDA

The GPU accelerated implementation of volume rendering using PyQt5 with modern OpenGL and NumPy.