The new, intelligent AI chip from SambaNova is capable of running models up to 5 trillion parameters, enabling fast and scalable inference and training, without sacrificing model accuracy.

SambaNova Systems, an AI chip startup, has introduced the SN40L chip, which enables customers to utilize more extensive AI models at a reduced cost. This chip is designed for running large language models in enterprise applications, accommodating models over twice the size of OpenAI’s ChatGPT.

SambaNova’s CEO, Rodrigo Liang, said the new AI chip was highly suitable for enterprise use cases, addressing considerations such as security, accuracy, and privacy that differ from consumer software like ChatGPT. This move is part of the growing competition in the AI chip market, where Nvidia (NASDAQ: NVDA), Intel (NASDAQ: INTC), Advanced Micro Devices (NASDAQ: AMD), and startups like SambaNova are vying for a share. (Reuters)

The SN40L chip supports 5 trillion parameter models and features advanced memory solutions to enhance AI data processing without compromising accuracy.

“SambaNova’s SN40L chip is unique. It addresses both HBM (High Bandwidth Memory) and DRAM from a single chip, enabling AI algorithms to choose the most appropriate memory for the task at hand, giving them direct access to far larger amounts of memory than can be achieved otherwise. Plus, by using SambaNova’s RDU (Reconfigurable Data Unit) architecture, the chips are designed to efficiently run sparse models using smarter compute,” said Peter Rutten, Research Vice-President, Performance-Intensive Computing, at IDC, in a statement.

The SN40L chip is manufactured by Taiwan Semiconductor Manufacturing Company (TPE: 2330) for SambaNova.

Related Story: SoftBank Vision 2 Leads $676M Round In SambaNova Systems

Image Credit: SambaNova Press Release