fasthardware 3.0.1

A low-level hardware optimization pipeline for zero-latency video streaming and multi-core acceleration.

fasthardware (v3.0.1)

A lightweight, low-level hardware optimization pipeline engineered to bypass Python's GIL bottlenecks, eliminate OpenCV camera buffer lag, and maximize multi-core execution pools (CPU/GPU) for computer vision architectures.

---

Key Features

• Lazy-Initialization Video Stream: Defeats console input() or engine compilation delays by spinning up cv2.VideoCapture only after blocking I/O operations are completely finished.
• Asynchronous Background Refreshes: Offloads video frame captures onto a dedicated background thread to guarantee zero-latency read() states.
• Multi-Backend Tuning: Direct runtime optimization hooks for NVIDIA CUDA/cuDNN, Intel OpenVINO, ONNX Runtime, and native OpenCV OpenCL pipelines.
• Low-Level Core Priority: Automatically elevates OS scheduling priority and forces high-performance thread-binding policies.

---

📦 Installation

pip install fasthardware