The rapid growth of artificial intelligence (AI) has created new challenges for data storage systems, which must handle vast amounts of data and provide high-performance access to support AI workloads. Cloudian, a company founded by Michael Tso '93, SM '93, and Hiroshi Ohta, has developed a scalable storage system that helps data flow seamlessly between storage and AI models.
Cloudian's solution reduces complexity by applying parallel computing to data storage and consolidating AI functions and data onto a single parallel-processing platform. This unified architecture eliminates complexity by integrating storage and AI inferencing into a single platform, reducing operational overhead and accelerating time-to-production for AI initiatives.
The system provides high-performance storage with industry-leading object storage read performance of 35GB/s per node, enabling faster AI model inference and improved application responsiveness. It also supports massive vector datasets while maintaining high-performance access for real-time inferencing workloads, making it ideal for applications like recommendation engines, computer vision, and natural language processing.
Cloudian's solution offers several benefits, including cost efficiency, flexibility, and scalability. By reducing the total cost of ownership compared to deploying separate storage and inferencing platforms, Cloudian's solution simplifies management and reduces data movement costs. It also supports both on-premises and hybrid cloud deployments, giving organizations maximum flexibility in their AI infrastructure strategy.
The effectiveness of Cloudian's technology has been demonstrated in various use cases, showcasing impressive results such as 11x faster genomic analysis throughput, 2.9M IOPS during peak inference, and $1.7M annual savings versus cloud alternatives. By providing a scalable and high-performance storage solution, Cloudian is helping businesses keep up with the demands of AI and unlock new insights from their data.
