SambaNova Systems offers RDU
Product
RDU is SambaNova's Reconfigurable Dataflow Unit, the company's AI chip family for inference and training at scale.
- Currently Offered
SambaNova's Reconfigurable Dataflow Unit (RDU) is a dataflow architecture chip family built for enterprise AI training and inference. The SN40L generation introduced a single-card design combining compute with high-capacity memory to run large models without external DRAM bottlenecks. The following SN50 generation continues the dataflow approach with larger on-chip memory and higher compute density for generative and agentic AI workloads.
SambaNova positions RDU-based systems for inference serving, including agentic workflows that require low latency across multiple model calls. The chips are packaged in the company's DataScale and Rack Scale systems, marketed as alternatives to GPU-centric AI infrastructure for private-cloud and on-premise deployments.
RDU is SambaNova's Reconfigurable Dataflow Unit, the company's AI chip family for inference and training at scale.
Analog compute-in-memory processor delivering 25 TOPS for edge AI inference.
Prometheus is a memory-first AI server that collapses racks of infrastructure into one unit.
Cerebras Wafer-Scale Engine is a family of wafer-scale AI processors.
Analog in-memory inference platform collocating deep sub-threshold compute inside multi-level non-volatile memory.
Anyscale provides a distributed computing platform for scaling AI and machine learning workloads.
World's largest AI chip: 4T transistors, 125 petaflops on one wafer.
Broadcom designs and supplies semiconductors for networking, wireless, broadband, and industrial markets.
Trainium is AWS's purpose-built AI accelerator chip family for training and inference of large generative models.
Tenstorrent PCIe accelerator card family for AI training and inference workloads.
Analog compute-in-memory processor delivering 25 TOPS for edge AI inference.
Prometheus is a memory-first AI server that collapses racks of infrastructure into one unit.
Cerebras Wafer-Scale Engine is a family of wafer-scale AI processors.
Analog in-memory inference platform collocating deep sub-threshold compute inside multi-level non-volatile memory.
Anyscale provides a distributed computing platform for scaling AI and machine learning workloads.
Sohu is a purpose-built inference chip designed for high-performance AI model serving.
World's largest AI chip: 4T transistors, 125 petaflops on one wafer.
Broadcom designs and supplies semiconductors for networking, wireless, broadband, and industrial markets.
TPU is Google's custom AI accelerator ASIC family delivered through Google Cloud TPU Pods for ML training and inference.
Trainium is AWS's purpose-built AI accelerator chip family for training and inference of large generative models.
Tenstorrent PCIe accelerator card family for AI training and inference workloads.
Intel Gaudi is a data-center AI accelerator product family.