Intel’s Habana Labs Launch Second-Generation Processors

Intel-acquired firm Habana Labs, announced this week the launch of its Habana Gaudi2 and Greco processors, its second-generation processors purposely built for deep learning applications that deliver twice the training throughput of chip rival Nvidia’s A100 GPU. 

These new processors provide customers with “high-performance, high-efficiency deep learning compute choices for both training workloads and inference deployments” while at the same time minimizing the AI barrier. In computational terms, the inference is the result of the capabilities learned during deep learning training. 

“Compared with the A100 GPU, implemented in the same process node and roughly the same die size, Gaudi2 delivers clear leadership training performance as demonstrated with apples-to-apples comparison on key workloads,” said Eitan Medina, chief operating officer at Habana Labs. “This deep-learning acceleration architecture is fundamentally more efficient and backed with a strong roadmap.” 

Founded in 2015, Habana Labs develops AI and deep learning computing solutions. The company is able to optimize efficiency and advance AI processing by offering separate processors for training and interference workloads in the data center. The company has lab locations in Caesarea and Tel Aviv in Israel, as well as India, Poland, and California. Habana Labs’ clientele and partners include autonomous driving technology developer Mobileye, American defense corporation Leidos, California-based information tech company Supermicro, and data storage company DDN. 

“The launch of Habana’s new deep learning processors is a prime example of Intel executing on its AI strategy to give customers a wide array of solution choices – from cloud to edge – addressing the growing number and complex nature of AI workloads,” said Sandra Rivera, Intel executive vice president and general manager of the data center and AI Group. “Gaudi2 can help Intel customers train increasingly large and complex deep learning workloads with speed and efficiency, and we’re anticipating the inference efficiencies that Greco will bring.”