In a significant development in the world of artificial intelligence, SambaNova, an AI chip startup that has raised over $1.1 billion in venture capital funding, has announced the launch of Samba-1, a new generative AI product aimed at enterprise customers. This move is seen as a direct challenge to OpenAI and other rivals in the AI industry.
Samba-1 is an AI-powered system designed for a variety of tasks such as text rewriting, coding, language translation, and more. The company describes the architecture as a “composition of experts,” which is essentially a bundle of 56 generative open source AI models.
Rodrigo Liang, the co-founder and CEO of SambaNova, claims that Samba-1 allows companies to fine-tune and address multiple AI use cases while avoiding the challenges of implementing AI systems ad hoc. He further stated that Samba-1 is fully modular, enabling companies to asynchronously add new models without eliminating their previous investment.
The main advantage of Samba-1, according to Liang, is that it is a collection of models trained independently rather than a single large model. This means that customers have control over how prompts and requests to it are routed. A request made to Samba-1 travels one of 56 directions, depending on the rules and policies a customer specifies, rather than through a single large model like GPT-4.
This multi-model strategy also reduces the cost of fine-tuning on a customer’s data, as customers only have to worry about fine-tuning individual or small groups of models rather than a massive model. Furthermore, it could result in more reliable responses to prompts, as answers from one model can be compared with the answers from the others, albeit at the cost of added compute.
Samba-1 can be deployed on-premises or in a hosted environment depending on a customer’s needs. While some vendors, including OpenAI, offer attractive pricing for fine-tuning large generative models, and several startups provide tools to route prompts among third-party models based on manually-programmed or automated rules, what SambaNova is offering is a full-stack solution with everything included, including AI chips, to build AI applications.
Liang stated that Samba-1 gives every enterprise their own custom GPT model, ‘privatized’ on their data and customized for their organization’s needs. The models are trained on customers’ private data, hosted on a single server rack, with one-tenth the cost of alternative solutions.
In conclusion, SambaNova’s Samba-1 is a promising new product in the AI industry, offering a set-it-and-forget-it package that includes everything needed to build AI applications. While it remains to be seen how it will compare to offerings from OpenAI and other rivals, its multi-model strategy and potential for cost savings and improved reliability make it an attractive option for enterprise customers.
Grow your business with AI. Be an AI expert at your company in 5 mins per week with this Free AI Newsletter