Independent AI chip maker Sambanova Systems, which has been slow to gain traction in the fast-growing AI market, launched a new chip and partnered with Intel to focus on the growing demand for agentic AI inference.
Sambanova on February 24 unveiled its latest AI chip, the SN50, which it claims is faster than competing chips. The hardware provider’s alliance with Intel — which reportedly considered acquiring Sambanova as recently as last year — includes providing high-performance, cost-efficient AI inference products.
In 2017 the startup said it also raised $350 million, though it did not disclose its market valuation. The vendor’s last funding round was in April 2021, when it raised $676 million in Series D.
The new chip has higher memory capacity and efficient”semioticsAccording to Sambanova, “Tokenomics – until recently, a term associated with blockchain technology – is the economics of token AI models used to process and generate data. The SN50 chip offers a software-optimized, reconfigurable dataflow processor that maps AI model graphs onto the chip to optimize inference.
The new chip comes amid a AI market Shifting our focus from general to agentic inference.
Inference is the stage of AI technology in which the model applies what it learns to new data. Agentic inference involves a multi-step reasoning process for agents to understand and act. This change in the type of estimation means that some hardware providers are emphasizing helping enterprises save money by doing more with less by using more cost-efficient chips. These speculation-focused sellers are attracting other mainstream sellers. For example, Cerebra’s partnership with OpenAI It’s a test bed to see how inference can be done without Nvidia GPUs and at a lower cost.
Sambanova’s opportunity
With Sambanova, in particular, is its emphasis on agentic AI inference at a time when the market is in demand and has the opportunity to showcase its technology.
“Sambanova has long established itself as a generic AI inference innovator, capable of running small models rapidly,” said Brendan Burke, an analyst at Futurum Group. He said that although the vendor still needs to find its ideal use case, agentic inference could be a differentiator for it.
“The use case of the agentic tool is optimal for Sambanova’s architecture“The recent growth of agents means this is the right time to scale up this technology, whereas previously, the market was really focused on larger models and did not value the ability to achieve quick performance on smaller models,” he said.
Despite this market window, Sambanova still faces some challenges. For one, the agentic inference market is still in its early stages, so Sambanova faces several strong competitors, notably Cerebras and Grok, as well as the dominant vendor in the AI chip market, Nvidia.
J. “There are a lot of players,” said Jack Gould, president of Gould Associates.
need to be unique
Additionally, vendors need to do more to differentiate, said Gartner analyst Gaurav Gupta.
“It’s much more than hardware performance that’s needed to gain traction,” Gupta said. He said enterprises also choose compute for its flexibility, integration and the vendor’s developer ecosystem.
However, Sambanova’s partnership with Intel could help.
“Intel has a big presence; it depends on how this all shapes up,” Gold said.
However, Intel also needs to find its ideal AI application, Burke said, noting that Intel has long been losing market share to AMD and Hand.
“This new wave open source agent The close alignment it provides with the portfolio encourages the development of agent-specific architectures, where AI accelerators and CPUs are combined to efficiently move data to each user, Burke said.
Being flexible is essential for enterprises keeping an eye on market movements, Gould said.
“In the next few years, we are going to see customized silos,” he said. “There will be change; it’s still a developing market. Don’t cast everything in cement.”
