SambaNova Reconfigurable Dataflow Unit (RDU)

peter_d_sherman2 pts0 comments

RDU | Next-Gen AI Chip for Inference at Scale

More results

En

Contact Us

Try It Now

Start Building

Contact Us

Products

SambaCloud

SambaStack

SambaManaged

Technology

SambaRack

RDU

Dataflow

Solutions

Inference Providers

Sovereign AI

Argyll (UK)

Infercom (EU)

OVHcloud (EU)

SouthernCrossAI (AU)

Developers & Enterprises

Government & Public Sector

Developers

Developer Showcase

Early Access Program

Community

Pricing

Docs

Resources

Case Studies

Blog

Events

Videos

White Papers

SambaAcademy

Support

Datacenter Walkthrough

Company

About Us

Team

Awards & Recognition

Newsroom

Press Releases

News Coverage

Careers

Contact Us

Contact Us

Start Building

Reconfigurable Dataflow Unit (RDU)

Delivering fast and energy-efficient inference

Introducing the SN50

Purpose-built for agentic inference, our fifth-generation chip, the SN50, is the only chip to deliver the speed and throughput required for agentic AI.

Built on the Dataflow Architecture, the SN50 delivers the best tokens per watt with 5X more compute and 4X more network bandwidth than our fourth-generation SN40.

Learn more

Headline here

From chips to racks

The combination of 16 SN40L RDUs creates a single, high-performance rack that can run the largest models, such as DeepSeek R1 671B and Llama 4 Maverick, with the fast inference. These racks can be seamlessly integrated into any existing air-cooled data center.

Learn more →

Seamlessly achieve high performance

From chips to racks

The combination of 16 SN40L RDUs creates a single, high-performance rack that can run the largest models, such as DeepSeek R1 671B and Llama 4 Maverick, with fast inference. These racks can be integrated seamlessly into any existing air-cooled data center.

Learn more →

Solving AI’s data movement problem

Data movement is the most expensive operation when running AI. SambaNova RDUs are designed to solve this problem by using an architecture that creates an assembly line process on the RDU chip.

Our Dataflow Architecture moves data seamlessly from operation to operation, saving power and time when processing the largest of models.

Learn more →

Tiered memory supports the largest models

RDUs employ a unique three-tier memory architecture that enables scaling to the largest LLMs. The infrastructure can scale to support running and switching between multiple models in milliseconds.

Learn more →

Find out how SambaNova is driving AI innovation

Contact Us

The best speed and throughput in the Goldilocks Zone

SambaNova RDUs deliver low latency with high throughput, resulting in better tokenomics for use cases like AI coding agents that require near-real time inference.

Learn more →

Energy-efficient AI inference

Our dataflow architecture delivers extraordinary performance without the overhead of moving data back and forth between memory like GPUs.

Our fourth-generation chip, the SN40, delivers fast inference with an average of just 10 kWh, allowing SambaRack systems to be air cooled.

Learn more →

From chips to racks

SambaNova RDUs combine to create a single platform that can run the largest models. The fifth-generation SN50 RDUs can scale up to 256 chips across multiple racks and run models that are up to 10 trillion parameters in size and with a context length of up to 10 million tokens.

With the RDU as the heart of SambaRack, these systems can be seamlessly integrated into existing air-cooled data centers.

Learn more →

Built for cloud scale

New with the SambaNova SN50 chips is a scaleout network of up to 32K RDUs. This enables huge cloud-scale inference services making it the ideal solution for inference service providers, like neo-cloud providers and hyperscalers.

Learn more →

Dataflow architecture

Our innovative compute and memory chip layout enables seamless dataflow between operations when processing AI models. This approach results in high-speed data traffic and significant gains in performance and efficiency.

Learn more →

Three-tier memory for efficiency

The SN40L design enables multiple models to run in memory and switch models in microseconds. This unique layout enables SambaNova to scale to the largest models, like DeepSeek and Llama 4 — all on a single rack.

Learn more →

Related resources

SambaNova Launches First Turnkey AI Inference Solution for Data Centers, Deployable in 90 Days

News -->

News

SambaNova Launches First Turnkey AI Inference Solution for Data Centers, Deployable in 90 Days

July 7, 2025

Why SambaNova's SN40L Chip Is the Best for Inference

Technology -->

Blog

Why SambaNova's SN40L Chip Is the Best for Inference

September 10, 2024

SambaNova Launches its AI Platform in AWS Marketplace

News -->

News

SambaNova Launches its AI Platform in AWS Marketplace

May 29, 2025

Choose the right RDU for your organization

Future-proof your infrastructure

Our fourth-generation RDU SN40 and fifth-generation SN50 are the heart of the SambaNova solution platform.

Speed

RDUs are the only solution...

inference sambanova learn models data rdus

Related Articles