Browse & filter

GPU-accelerated password recovery and hash cracking supporting hundreds of algorithms and attack modes.

High-throughput LLM serving with PagedAttention, continuous batching, and OpenAI-compatible APIs for GPU clusters.

Structured generation language for fast serving: RadixAttention, constrained decoding, and multi-turn batching for frontier-class workloads.

NVIDIA TensorRT–based library for optimized LLM inference on GPUs with multi-GPU and speculative decoding features.

Composable transformations (grad, vmap, pmap) plus NumPy-like API for high-performance ML research on accelerators.

Multi-framework inference server for TensorRT, ONNX, PyTorch, Python backends—dynamic batching, ensembles, and GPU sharing.

Alibaba’s high-performance LLM inference engine (CUDA-focused) for production serving of diverse decoder architectures.

Physics-ML / scientific deep learning framework: neural operators, PINNs, and domain-parallel training on GPUs.

GPU-accelerated cross-platform terminal emulator.

GPU terminal with ligatures, images, and multiplexing features.

Redirects OpenGL rendering to a server GPU and streams frames for remote 3D and video workloads.