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 develops both hardware and software systems aimed at improving the efficiency and speed of AI model training and inference, serving enterprises and research institutions.
Founded in November 2017, SambaNova emerged from research on microprocessing arrays for machine learning, with a focus on reconfigurable dataflow architecture. The company has raised significant venture capital and has deployed its systems in cloud and on-premises environments, including partnerships with major computing centers. Its technology targets the growing demand for specialized infrastructure 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 Stanford University professor known for work in parallel computing, and Ré having a background in machine learning systems.
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, as well as expansion into cloud-based services. In the early 2020s, the company started offering cloud-based AI services and inference platforms, allowing customers to access its accelerators without owning physical hardware.
In 2023, SambaNova introduced the SN40L reconfigurable dataflow unit (RDU), a processor designed for AI models and inference workloads. This chip represented a generational step in the company's hardware roadmap, aiming to handle increasingly complex large language models and generative AI applications. In 2024, Time listed SambaNova Suite on its annual Best Inventions list, and in 2026, the company was included in the Forbes AI 50, recognizing its influence 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 such as GPUs, which use fixed instruction sets, the RDU is designed to be reconfigured at runtime to match the dataflow patterns of specific AI models. This approach aims to reduce data movement bottlenecks and improve computational efficiency for workloads like deep learning and neural networks.
The company's hardware platform is built around the RDU, a processor architecture intended to improve data movement and computation for AI workloads. SambaNova offers both cloud and on-premises systems used for AI inference and model deployment. Its software stack, including the SambaFlow environment, allows developers to compile models from frameworks like PyTorch and TensorFlow to run on the RDU, abstracting away low-level programming complexities.
SambaNova's systems are positioned as an alternative to mainstream accelerators from companies like Nvidia (not listed) and AMD, with a focus on memory bandwidth and reconfigurability. The SN40L RDU, introduced in 2023, incorporates on-chip memory and high-bandwidth interfaces to support large model sizes and high-throughput inference. The company also provides tools for model optimization, including quantization and pruning, to maximize performance on its hardware.
Deployments and Partnerships
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 deployment allowed researchers to evaluate the RDU's performance for scientific AI workloads, such as climate modeling and materials science.
In Japan, the Riken Center for Computational Science (R-CCS) deployed SambaNova systems to operate alongside the Fugaku supercomputer, one of the world's fastest supercomputers. This integration aimed to explore hybrid computing approaches, combining traditional high-performance computing with AI-specific accelerators. SambaNova's technology is also used in newer AI inference projects, including Vector Core Compute and OVHcloud to power their AI Endpoints platform, providing cloud-based access to SambaNova hardware for enterprise customers.
These deployments highlight SambaNova's strategy of targeting both research institutions and commercial cloud providers. By partnering with established computing centers, the company gains credibility and real-world validation for its architecture, while cloud partnerships expand its reach to smaller organizations that prefer managed AI services.
Market Position and Competition
SambaNova operates in the competitive AI hardware market, facing rivals such as Cerebras Systems and Groq, which also develop specialized accelerators for AI workloads. Unlike these competitors, SambaNova emphasizes its reconfigurable dataflow architecture as a differentiator, claiming flexibility across various model types and sizes. The company's valuation of $5.1 billion as of April 2021 placed it among the most valuable AI hardware startups, though it has not disclosed recent funding rounds.
The company's focus on inference, rather than just training, aligns with the growing deployment of AI applications in production environments. As large language models become more widespread, demand for efficient inference hardware is expected to rise. SambaNova's cloud services, such as SambaNova Suite, aim to capture this market by offering a full-stack solution, from hardware to software, with a pay-as-you-go model.
However, the company faces challenges from established players like Nvidia (not listed) and Intel, which have deep resources and broad software ecosystems. SambaNova's success depends on its ability to demonstrate superior performance and cost-effectiveness for specific AI workloads, as well as building a robust developer community around its software stack.
Future Directions
Looking ahead, SambaNova is likely to continue iterating on its RDU architecture, with potential improvements in memory capacity, interconnect speed, and energy efficiency. The company may also expand its software offerings to support a wider range of AI frameworks and models, including emerging architectures like transformers and diffusion models (not listed). As of 2026, SambaNova's inclusion in the Forbes AI 50 suggests sustained relevance in the AI industry, though specific product roadmaps have not been publicly detailed.
The company's partnerships with research institutions and cloud providers position it to benefit from government and enterprise investments in AI infrastructure. With the increasing focus on generative AI across sectors, SambaNova's hardware could play a role in enabling cost-effective inference at scale. However, the rapidly evolving AI hardware landscape means the company must continuously innovate to maintain its competitive edge.