Download Area

Home > Machine Learning

nGraph (free) Download Full | **UPDATE

- nGraph

nGraph (free) Download Full | **UPDATE

Published Date: 2024-04-14

nGraph Free Download

nGraph is an open-source library that contains tools to optimise and execute a TensorFlow or PyTorch model on a variety of backends. The backends include CPUs, Intel's Neural Compute Stick 2, ARM CPUs, MyriadX VPUs and GPUs. nGraph has a user-friendly interface, improved performance, extended capabilities and robust production. Also, it supports deep learning applications by computer vision and natural language processing.

nGraph is crucial for developers and researchers to optimise models for deployment across various platforms. It offers a comprehensive set of optimisations, including operator fusion, constant folding, and layout optimisations. nGraph also allows custom layers and operations to be easily integrated into the optimisation process. Furthermore, it provides a rich set of tools for debugging and profiling models, making it easier to identify and resolve performance bottlenecks. To download nGraph, visit the official website and select the appropriate version for your operating system and hardware.


nGraph: Frameworks using nGraph Compiler stack to execute workloads have shown up to 45X performance boost when compared to native framework implementations. We've also seen performance boosts running workloads that are not included on the list of Validated workloads, thanks to nGraph's powerful subgraph pattern matching. Additionally, we have integrated nGraph with PlaidML to provide deep learning performance acceleration on Intel, nVidia, & AMD GPUs. nGraph Compiler aims to accelerate developing AI workloads using any deep learning framework and deploying to a variety of hardware targets. We strongly believe in providing freedom, performance, and ease of use to AI developers. Our documentation has extensive information about how to use nGraph Compiler stack to create an nGraph computational graph, integrate custom frameworks, and to interact with supported backends.