Nod Labs

About Nod Labs

Nod.ai, initially founded as Nod Labs in 2013 by Anush Elangovan and Harsh Menon, has pivoted from its early focus on hardware to become a key player in artificial intelligence software. The company's original mission centered on gesture recognition and motion-tracking wearables, such as rings for virtual reality, backed by investors like Atlantic Bridge, Sequoia Scout Seed Fund, and Walden Riverwood Ventures. Co-founder Anush Elangovan brought experience from Google, where he worked on Chromebooks, and Cisco, while Harsh Menon was an early employee at Zee.Aero/KittyHawk, working on flying cars.

The company transitioned to what is now Nod.ai, concentrating on open-source AI software that optimizes the performance of machine learning models. This strategic shift culminated in its acquisition by AMD in October 2023 for an undisclosed amount. The acquisition was a strategic move by AMD to bolster its open software ecosystem and better compete in the AI hardware and software market by integrating Nod.ai's expertise.

Nod.ai's core business revolves around advanced compiler and code-generation technologies that enable customers to train and deploy machine learning models with greater efficiency. The firm's flagship product, the SHARK software suite, is a compiler-based automation tool that reduces the need for manual optimization when deploying AI models across various hardware platforms. This technology is designed to make AI models run faster and more efficiently, a critical factor for clients in cloud computing, data centers, and edge computing. Nod.ai is a major contributor to important open-source AI projects, including Torch-MLIR and OpenXLA/IREE, which are foundational technologies for compiling and running AI programs.

By creating solutions that tune AI models to run optimally on specific hardware, Nod.ai addresses a crucial bottleneck in the deployment of AI. Its software accelerates AI solutions for a wide range of AMD hardware, including Instinct data center accelerators, Ryzen AI processors, and Radeon GPUs. This focus on compiler technology provides a significant performance lift, enabling models to run with lower latency across diverse platforms, from large-scale cloud servers to endpoint devices.