The infographics shows AI GPU cluster with scaling of compute and storage designed for AI workloads. Some of the key configuration highlights include:

Skylus.ai appliance provides immense scalability attaching high-performance storage as per the workflow requirement.
The system also includes Tyrone GPU Servers with Skylus.ai for compute heavy task such as AI training and simulations, Tyrone ParallelStor Metadata Nodes to manage file system metadata efficiently, Tyrone ParallelStor Data Nodes to handle large scale data storage. The below architecture supports separate scaling of compute and storage, allowing flexibility to add GPU servers or storage nodes independently based on workload needs.

Research as a Service – Solution Stack

Research as a Service – Solution Stack – Tier 0 Storage

Building Blocks Specifications