NervanaSystems/ngraph

Health v2 · maintenance only

Flags: archived

Adoption not part of the score

1344 stars · 218 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

nGraph is a C++ deep learning graph compiler that accelerates neural network workloads from frameworks like TensorFlow, ONNX, MXNet, and PyTorch across diverse hardware targets. The project has been merged into Intel's OpenVINO toolkit and is no longer maintained here.

Use cases

• compile deep learning graphs for faster inference
• accelerate tensorflow workloads on intel hardware
• run onnx models with a graph compiler
• optimize neural network subgraphs for performance
• deploy deep learning models to intel npu accelerators

When to choose

• you need the historical ngraph compiler stack for research or legacy builds
• you want to study graph compiler architecture for deep learning

When to avoid

• you want active maintenance or new features - use OpenVINO instead
• you need current framework version support
• you need GPU support via PlaidML going forward

Facets

library · maturity abandoned

compiler machine-learning deep-learning llm-training deep-learning machine-learning compilers performance cpp python neural-network-compiler graph-compiler onnx tensorflow-integration moved-to-openvino hardware-acceleration linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="NervanaSystems/ngraph")