MinIO Deepens NVIDIA Integrations to Deliver Ultra-Efficient, Low-Cost AI Infrastructure

MinIO, the leader in high-performance object storage for AI, is announcing that it has deepened its support for the NVIDIA AI ecosystem with new advancements to MinIO AIStor—the object store designed for the exascale data infrastructure challenges associated with modern AI workloads. These latest integrations will improve the efficiency of AI infrastructures while maximizing utilization and strealiming management, allowing enterprise teams to focus on more strategic AI initiatives, according to MinIO. 

MinIO AIStor now features the following:

• Support for NVIDIA GPUDirect Storage (GDS) for object storage: This innovation frees up compute for additional AI data processing while reducing infrastructure costs, increasing CPU efficiency on the NVIDIA GPU server. By leveraging Ethernet fabrics, MinIO AIStor's GDS for object storage implementation establishes a direct path between MinIO AIStor and NVIDIA GPU memory, improving GPU server efficiency and scaling adaptability. 

• Native integration with NVIDIA BlueField-3 networking platform: As the first and only object storage software to run natively on NVIDIA’s BlueField-3 Data Processing Unit (DPU), MinIO AIStor’s integration helps lower object storage TCO without sacrificing performance, according to MinIO. This ensures that AI workloads operating at petabyte to exabyte scale are optimized with an ultra-efficient, low-cost architecture.

• Incorporation of NVIDIA NIM microservices into AIStor promptObject inference: Now, MinIO AIStor simplifies AI-powered interactions with stored objects through the incorporation of AIStore promptObject into NVIDIA NIM microservices inference infrastructure.This not only simplifies deployment and management of inference instructure, but allows users to “talk” to unstructured objects—in the same way one would engage an LLM—to deliver faster inference via model optimizations for NVIDIA hardware. 

"MinIO's strong alignment with NVIDIA allows us to rapidly innovate AI storage at multi-exabyte scale, leveraging their latest infrastructure," said AB Periasamy, co-founder and co-CEO, MinIO. "This approach delivers high-performance object storage on commodity hardware, enabling enterprises to future-proof their AI, maximize GPU utilization, and lower costs."

“We are excited to see MinIO bring AIStor to the NVIDIA AI ecosystem and to explore how AIStor and GPUDirect Storage together perform under the specific demands of our workloads,” said Alex Timofeyev, director, high performance compute engineering and operations, Recursion. “Our work requires high scalability, throughput, and efficiency in handling AI workloads. Based on preliminary testing, we believe that MinIO AIStor will increase efficiency in the CPU's on our AI compute infrastructure, ultimately enhancing the performance and economics in our data environment.” 

To learn more about MinIO’s latest enhancements with NVIDIA, please visit https://min.io/.