SambaNova Systems, Inc. is an American artificial intelligence (AI) and semiconductor hardware company that designs and manufactures hardware accelerators termed Reconfigurable Dataflow Units (RDUs), adjusted for deep learning models and generative AI applications. The company focuses on providing both cloud-based and on-premises computing systems for AI inference and model deployment, aiming to improve data movement and computation efficiency for AI workloads.
Founded in November 2017, SambaNova has grown into a significant player in the AI hardware space, raising substantial funding and deploying its technology in research laboratories and commercial cloud platforms. Its architecture, based on reconfigurable dataflow, distinguishes it from traditional GPU-based approaches, offering an alternative for organizations seeking to run large-scale AI models.
History
SambaNova was co-founded in November 2017 by Kunle Olukotun, Christopher Ré, and Rodrigo Liang. The technical foundation of the entity was derived from microprocessing array research for machine learning systems that received funding from the Defense Advanced Research Projects Agency (DARPA). The founders brought expertise from academia and industry, with Olukotun being a professor at Stanford University and Ré a professor at the same institution, while Liang had industry experience.
Between 2018 and 2021, SambaNova raised about $1.1 billion, reaching a valuation of $5.1 billion by April 2021. This funding supported the development of its hardware and software stack. In the early 2020s, the company started cloud-based AI services and inference platforms, expanding its offerings beyond hardware sales.
In 2023, SambaNova introduced the SN40L reconfigurable dataflow unit (RDU), a processor designed for AI models and inference workloads. This chip represented a significant step in the company's technology roadmap, targeting the growing demand for generative AI inference.
In 2024, Time listed SambaNova Suite on its annual Best Inventions list, recognizing the company's software platform. In 2026, it was included in the Forbes AI 50, highlighting its continued relevance in the AI industry.
Technology
SambaNova develops AI computing systems based on a reconfigurable dataflow architecture intended for machine learning and generative AI applications. Unlike traditional processors that execute instructions sequentially, the RDU is designed to configure its data paths dynamically, optimizing the flow of data for specific AI workloads. This approach aims to reduce the overhead of data movement, which is often a bottleneck in AI computations.
The hardware platform is built around the Reconfigurable Dataflow Unit (RDU), a processor architecture intended to improve data movement and computation for AI workloads. The RDU can be programmed to support various neural network models, including large language models and other deep learning architectures.
The company offers cloud and on-premises systems used for AI inference and model deployment. Its software stack, SambaNova Suite, provides tools for model training, fine-tuning, and inference, aiming to simplify the deployment of AI applications.
Research and Supercomputing
SambaNova computing hardware platforms have been used in public research labs and supercomputing complexes to test alternative acceleration models for scientific calculations. The United States Department of Energy's Argonne National Laboratory integrated the SambaNova computing framework within its specialized AI Testbed located at the Argonne Leadership Computing Facility (ALCF). This integration allowed researchers to explore the RDU's capabilities for scientific workloads.
In Japan, the Riken Center for Computational Science (R-CCS) deployed SambaNova systems to operate alongside the Fugaku supercomputer. This deployment aimed to evaluate the RDU's performance in a high-performance computing environment, potentially complementing traditional CPU and GPU resources.
These collaborations demonstrate SambaNova's focus on the research community, providing an alternative to mainstream accelerators for AI-driven scientific discovery.
Commercial Deployments
SambaNova's technology is also used in newer AI inference projects, including Vector Core Compute and OVHcloud to power their AI Endpoints platform. These partnerships allow cloud providers to offer SambaNova-based inference services to their customers, expanding the reach of the RDU architecture.
The company's commercial strategy involves both direct sales to enterprises and partnerships with cloud service providers, enabling a broader adoption of its hardware without requiring customers to invest in on-premises infrastructure.
Funding and Growth
SambaNova has attracted significant investment from venture capital firms and strategic investors. The funding raised between 2018 and 2021, totaling about $1.1 billion, supported the company's research and development efforts, as well as the expansion of its cloud services. The valuation of $5.1 billion as of April 2021 positioned SambaNova among the most valuable AI hardware startups.
The company's growth has been driven by the increasing demand for AI inference, particularly for large language models and other generative AI applications. As organizations seek efficient ways to run these models, SambaNova's reconfigurable architecture offers a potential advantage in terms of performance and energy efficiency.
Competitive Landscape
SambaNova operates in a competitive market that includes established players like NVIDIA and AMD, as well as other startups such as Groq and Graphcore. While NVIDIA's GPUs dominate the AI training market, SambaNova differentiates itself by focusing on inference and its reconfigurable dataflow architecture, which can be tailored to specific models.
The company's approach is often compared to that of Groq, which also develops AI accelerators with a focus on inference. However, SambaNova's RDU is more flexible, allowing for reprogramming to support different model architectures, whereas Groq's chip is more specialized.
Future Directions
Looking ahead, SambaNova aims to continue improving its RDU architecture and software stack to keep pace with the evolving requirements of AI models. The company is likely to expand its cloud offerings and forge additional partnerships with research institutions and cloud providers.
As of 2024, SambaNova has not publicly disclosed specific plans for future chip generations, but the recognition from Time and Forbes suggests that its products are gaining traction. The company's ability to innovate in the rapidly changing AI hardware landscape will be crucial for its long-term success.