Inte is strengthening its datacenter AI plans with the acquisition of Israeli chip startup Habana Labs for $2B.

Habana Labs has been one of the first to deliver working hardware with impressive performance claims for both training and inference processing. Habana Labs launched its Goya chip for inference processing in September 2018, claiming roughly 3X performance advantage over NVIDIA with lower latency. The company subsequently announced its training chip, Gaudi in June 2019, claiming record performance and an integrated fabric based on industry standards to enable scaling to process very large AI models.

For its part, Intel has recently reaffirmed its plans to deliver its Nervana chips. Although it is not clear where the Nervana chips fit into the Habana's architecture, the Nervana software stack could.

The Habana network fabric could also one of the key reasons that Intel decided to abandon Nervana in favor of Habana’s technology. Nervana’s Neural Network Processor (NNP-T) uses a proprietary interconnect for scaling, while Habana’s Gaudi can scale to thousands of nodes over standard 100Gb Ethernet. And Gaudi even supports Remote Direct Memory Access, RDMA, which enables software to access memory across the fabric without taxing the remote CPU. This fabric can dramatically increase the performance of training very large neural network models, which are doubling in size every 3.5 months to handle ever-more complex AI tasks.

In a similar vein, NVIDIA is acquiring Mellanox for its networking technology to interconnect NVIDIA GPU’s.

“This acquisition advances our AI strategy, which is to provide customers with solutions to fit every performance need – from the intelligent edge to the data center,” said Navin Shenoy, executive vice president and general manager of the Data Platforms Group at Intel. “More specifically, Habana turbo-charges our AI offerings for the data center with a high-performance training processor family and a standards-based programming environment to address evolving AI workloads.”

Habana will remain an independent business unit and will continue to be led by its current management team. Habana will report to Intel’s Data Platforms Group, home to Intel’s portfolio of data center class AI technologies. This combination gives Habana access to Intel AI capabilities, including resources built over the last three years with deep expertise in AI software, algorithms and research that will help Habana scale and accelerate.

Habana chairman Avigdor Willenz has agreed to serve as a senior adviser to the business unit as well as to Intel. Habana will continue to be based in Israel.