WEKA Introduces High-Performance AI Data Platform Appliance For NVIDIA Systems

WekaIO, a leading AI-native data platform company, has recently unveiled its latest innovation, the WEKApod Data Platform Appliance. This new appliance is certified for NVIDIA DGX SuperPOD with NVIDIA DGX H100 systems, marking a significant milestone in the field of AI applications. The WEKApod combines WEKA's sophisticated AI-native data platform software with state-of-the-art storage hardware, offering a comprehensive solution ready for immediate deployment in AI projects.

The integration of the WEKApod Data Platform Appliance into the NVIDIA DGX SuperPOD solution empowers customers to swiftly provision compute, storage, and networking resources. This capability is crucial for accelerating the deployment of AI applications. The appliance boasts an impressive performance, with a single cluster capable of delivering up to 18,300,000 IOPs. It also excels in write performance for AI model checkpointing, scalability, resiliency, and support for real-time applications.

The WEKA Data Platform's AI-native architecture is designed to provide fast AI storage and high performance for AI data pipelines. It ensures low latency across all file sizes and offers high levels of write throughput for checkpointing operations. Additionally, the platform is engineered to maximize energy efficiency. By enhancing performance density, cooling efficiency, power distribution, and reducing idle energy consumption, customers can significantly decrease their carbon footprint—by up to 260 tons of CO2e per petabyte stored.

Moreover, the WEKA Data Platform facilitates efficient storage access. This efficiency ensures that computational tasks such as training and inference processes are executed with maximum performance while minimizing energy consumption. Consequently, organizations can lower their energy costs and carbon footprint of AI deployments, aligning with sustainability and environmental objectives.

WEKA offers a variety of deployment options for its Data Platform software, including on-premises, hyperscaler, hybrid, and GPU clouds. The company stands out as the first to provide a subscription software solution compatible with hardware from leading server vendors and major hyperscale public clouds.

The certification of the WEKApod Data Platform Appliance for NVIDIA DGX SuperPOD is a game-changer for enterprise AI customers. It provides a robust data infrastructure foundation that supports quick and efficient implementation of AI projects. This certification enables organizations to address complex AI workloads at data center scale and expedite their AI transformation journey.

Charlie Boyle, Vice President of DGX Systems at NVIDIA, highlighted the growing demand for high-performance infrastructure to support enterprise AI adoption. "The WEKApod Data Platform Appliance certification for NVIDIA DGX SuperPOD gives enterprises a proven solution at data center scale for tackling the most complex AI workloads, powering their AI transformation," Boyle stated.

WekaIO's commitment to overcoming data challenges and accelerating discoveries and insights has positioned the company as a visionary in its field. It has been recognized as a Visionary for three consecutive years in the Gartner Magic Quadrant for Distributed File Systems and Object Storage. With its innovative solutions like the WEKApod Data Platform Appliance, WekaIO continues to support leading enterprises and research organizations in their quest to harness the full potential of AI.