SambaNova, the AI chip company delivering the highest performance inference, today announced a strategic collaboration with Supermicro. The two companies will build and sell AI inference racks that combine SambaNova SN50 RDU chips with GPUs. Initially, the plan covers 10,000 SN50 RDU chips.
Supermicro is an AI inference racks, Enterprise, Storage, and 5G/Edge Total IT Solution Provider. Moreover, it offers Data Center Building Block Solutions® (DCBBS). The new SN50 inference racks will operate alongside NVIDIA GPUs as part of that DCBBS portfolio. In addition, the systems target disaggregated inference, which splits the workload across different processors.
“Supermicro is committed to delivering flexible, high-performance AI infrastructure that supports the evolving requirements of our customers. Our collaboration with SambaNova expands our broad DCBBS portfolio by enabling heterogeneous architectures that integrate GPU computing with specialized inference acceleration, helping customers deploy efficient, scalable AI solutions,” said Charles Liang, president and CEO of Supermicro.
“Agentic AI is the next wave of inference, and it has to run everywhere,” said Rodrigo Liang, co-founder and CEO of SambaNova. “Supermicro builds AI infrastructure for data centers globally. Now customers can order SN50 RDUs alongside GPUs, from a world-leading partner they trust. That’s how fast, efficient inference gets to scale.”
Supermicro to Build, Sell and Support the Systems
Agents work nonstop, and their speed depends on how fast the model responds. Therefore, Supermicro’s systems pair NVIDIA GPUs with SambaNova SN50 chips. The GPUs read the prompt. Meanwhile, the SN50 chips write the response quickly.
Supermicro will integrate, test, deliver, and support the systems for customers. That support includes on-site service. Meanwhile, SambaNova engineers will get models running and tuned on SN50. They will also help Supermicro with deployments. The companies will start with NeoClouds and sovereign clouds. Afterward, they will expand to enterprises that run AI in their own data centers.
How SN50 RDU Chips Work Alongside NVIDIA GPUs
SambaNova’s AI inference racks SN50 is the company’s fifth-generation Reconfigurable Dataflow Unit (RDU) chip. Notably, it is purpose-built for high-performance inference. RDUs serve the largest models at high speed and high throughput. Furthermore, they do not require extreme rack power density. Consequently, operators can deploy them far more easily in existing data centers.
RDUs also work alongside existing NVIDIA GPUs to drastically improve system performance and efficiency. In this setup, the GPUs run the prefill phase, while the RDUs run the decode phase. Therefore, operators can add inference capacity without waiting for next-generation data center build-outs. Power and cooling constrain those build-outs. Additionally, operators can extend the useful life of the GPUs they already own.
As a result, SambaNova and Supermicro can deliver substantially more high-performance inference capacity in 2027 than would otherwise be possible. Ultimately, these AI inference racks give customers a faster path to scalable, efficient AI.
Explore IT Tech News for the latest advancements in Information Technology & insightful updates from industry experts!
News Source: Businesswire.com
