Sunday, March 3, 2024

#MWC24: Marvell on Implementing AI Directly on the RAN

New possibilities are emerging now that Marvell is open sourcing software for building machine learning models that can be executed on OCTEON 10 processors, which are widely deployed in base stations around the world.

Joel Brand, SR Director of Product Marketing from Marvell, explains:

- Marvell's Octeon product line, used by industry giants like Nokia and Samsung, has embedded an AI accelerator, or NPU, to leverage machine learning and AI in the Radio Access Network.

- The unique implementation of this NPU is exposed through MLDEV, part of the Linux Foundation's Data Plane Development Kit (DPDK) , into the broader development community via the Apache TVM project, which is an open-source machine learning (ML) compiler framework for accelerating the deployment of deep learning models on various hardware platforms. 

- This innovation allows the use of standard open-source tools to configure and innovate directly on the Radio Access Network, potentially driving incremental revenue in the Telco environment.


https://youtu.be/sXQJJuNP2WQ

Like our videos? Want to participate? Contact info@nextgeninfra.io