Purpose-Built for the Most Demanding AI and Deep Learning Workloads of Today and Tomorrow
SambaNova DataScale® delivers the industry’s only comprehensive software-defined hardware solution to power the most advanced deep learning workloads. The combination of state-of-the-art hardware and a full stack software solution provides organizations with the ability to run the largest AI models and most demanding workloads with extreme performance, unmatched ease of use, and massive scalability.
EASE OF USE
Use common open standards and AI frameworks with no CUDA programming
Seamlessly scale to hundreds of systems with little to no code changes
Achieve the highest performance and efficiency for advanced AI workloads
Benefit from record setting accuracy on the world’s largest data sets and models
Unrivaled Ease of Use
Designed with a software first approach, DataScale allows organizations to use common frameworks, such as PyTorch. The compiler requires little to no code change, so they can break free from legacy systems that add unnecessary complexity by requiring the use of proprietary coding languages, such as CUDA.
Massive Scalability, Simplified
Designed to scale seamlessly, DataScale can grow to hundreds of systems to easily meet the increasing demands of advanced AI, deep learning, and state-of-the-art computing environments. DataScale overcomes the limitations of GPU-based systems which require significant work to install, configure, and manage the infrastructure.
DataScale automatically manages the flow of data to simplify scaling across any number of systems.
Delivering Next Generation Performance Today
Break free from the limitations of the legacy, instruction set architecture (ISA) with a platform that is powered by a next generation processor, designed with a dataflow architecture and purpose built for the largest deep learning models in the world.
DataScale takes advantage of the SambaFlow™ software stack to automatically configure the SambaNova Reconfigurable Dataflow Unit (RDU) to process your data with maximum efficiency in real time. The RDU is designed to let data flow through the processor, reducing the context switching and excess data movement found in CPU and GPU based designs. The SambaNova RDU powers DataScale to deliver the performance you need for today’s demanding workloads with the scalability to handle the needs of tomorrow.
Record Setting Accuracy on the Largest Models and Data Sets
Organizations benefit from the unprecedented accuracy that can be achieved with SambaNova. Leading data scientists leverage DataScale’s performance and ability to handle large volumes of data to deliver record setting accuracy, while powering the largest models in the world, on a single unified platform.
The core infrastructure to develop and deploy next-generation AI algorithms and applications at scale with unmatched performance, state-of-the-art accuracy, scale and ease of use.
Take a 3D Tour of DataScale
Peek inside our technology with this interactive tool.
Hear What Our Customers Say
Lawrence Livermore National Laboratory
“The support for running multiple models built into the hardware was a significant aspect of our selection of SambaNova Systems. … We’re seeing roughly a 5X improvement compared toa comparable GPU running the same models.”
Bronis de Supinski, Chief Technology Officer
Argonne National Laboratory
“We’re interested in technology … that can train models faster than individual GPUs, that has more scalability and hasmore performance, and so the SambaNova DataScale system met that criteria.”
Rick Stevens, Associate Laboratory Director
Hugging Face Demo
Hugging Face is the most popular model repository for developing transformer-based deep learning apps. Integrating Hugging Face with DataScale you can run thousands of Transformer models with state-of-the-art accuracy in seconds and with zero code changes.
Schedule a Meeting
Learn how SambaNova can advance your AI initiatives to help you achieve your impossible.