At the GTC 2025 Conference, Supermicro, Inc. (NASDAQ: SMCI), a prominent player in the IT solutions sector, announced the launch of a groundbreaking Petascale All-Flash Storage Server. This innovative server utilizes the NVIDIA Grace CPU Superchip, specifically designed to meet the high-performance demands of software-defined AI storage workloads. Supermicro's latest offering exemplifies its commitment to delivering optimized storage solutions tailored for artificial intelligence (AI), machine learning (ML), analytics, and enterprise storage environments.

2. Key Features of the New Storage Server

Supermicro's new storage server, built on its Building Block Solutions® architecture, offers significant advancements in performance and efficiency. The ARS-121L-NE316R 1U storage system incorporates a symmetrical design ideal for scaling AI workloads. Some of the standout features include:

Enhanced Processing Power: Powered by the NVIDIA Grace CPU, which includes 144 Arm Neoverse V2 cores, the server achieves high-performance I/O capabilities crucial for software-defined storage tasks.

Massive Storage Capacity: The system supports up to 16 hot-swap EDSFF PCIe Gen5 E3.S NVMe drives, allowing for a staggering 983TB of raw capacity using 61.44TB SSDs. A single rack can accommodate 40 systems, resulting in a total raw storage capacity of 39.3PB.

Energy Efficiency: The integration of the NVIDIA Grace CPU Superchip is designed to provide exceptional performance while consuming less power compared to traditional x86 servers.

3. Collaboration with Industry Leaders

Supermicro partnered closely with NVIDIA and WEKA to refine this storage solution, ensuring compatibility with a range of software applications. Rob Davis, Vice President of Storage Networking Technology at NVIDIA, commented on the collaboration, stating, "This new storage server allows customers to benefit from NVIDIA's innovations in both CPUs and DPUs to accelerate networking protocols, such as GPUDirect Storage, while using less power than comparable x86 servers."

WEKA, renowned for its high-performance storage software used in AI, high-performance computing (HPC), and enterprise data management, is also collaborating on this initiative. Nilesh Patel, Chief Product Officer at WEKA, highlighted the performance advantages observed during initial testing, noting significant scalability and power efficiency when using WEKA's Data Platform software on Supermicro's new storage server.

4. The Future of AI Storage Solutions

Charles Liang, President and CEO of Supermicro, emphasized the company's focus on innovation and customer choice, stating, "We have developed the Petascale storage server using the NVIDIA Grace CPU... Supermicro continues to bring to market the most advanced and optimized storage solutions available." This sentiment underscores Supermicro's commitment to leading the charge in next-generation storage technologies.

5. Presentation at GTC 2025

The new Petascale storage server will be featured in greater detail during an On-Demand session at NVIDIA GTC in San Jose, which commenced on March 17. This presentation aims to provide attendees with an in-depth understanding of the server's capabilities and its potential impact on the future of AI and ML storage solutions.

6. Conclusion

Supermicro's introduction of the Petascale All-Flash Storage Server marks a significant advancement in the field of high-performance storage solutions. By leveraging the NVIDIA Grace CPU Superchip and collaborating with industry leaders, Supermicro is poised to redefine the landscape of software-defined storage, making it an essential player for enterprises looking to optimize their AI and ML workloads. As the demand for efficient, scalable, and powerful storage systems continues to grow, Supermicro's innovative offerings will undoubtedly play a pivotal role in shaping the future of IT infrastructure.

Original press release