A leading university in Singapore has boosted the speed of its AI research with Xinnor’s xiRAID storage technology, according to a case study [PDF] published by the vendor.
To support its ongoing research, the unnamed university implemented multiple GPUs and compute systems, including hardware from Nvidia.
The university’s deployment supports advanced AI projects in fields including healthcare and natural language processing, and features a BeeGFS cluster of clients and servers. BeeGFS is a parallel file system developed for high-performance computing, and includes a distributed metadata architecture for scalability and flexibility.
For its storage requirements, the university required fast access to data, data protection, ease of deployment and management, cost-optimization, and scalability in response to future growth in terms of clients and performance requirements.
To deliver on these requirements, a storage deployment – architected by local Xinnor partner On Demand System (ODS) – is delivering a sequential read performance of 24.7Gbit/sec through two 100Gb InfiniBand ports.
The infrastructure includes two server nodes equipped with 24 NVMe drives each, protected and accelerated by xiRAID. The configuration delivers not only the required throughput, but also ensures data protection and scalability for future growth.
The software-driven xiRAID technology employs advanced algorithms to provide redundancy and fault tolerance, ensuring that data remains safe even in the event of multiple drive failures. For future performance requirements, the university can add more InfiniBand cards.
“xiRAID’s ability to achieve near-perfect resource utilization and maximize NVMe performance is a game-changer for academic research,” declared Davide Villa, chief revenue officer of Xinnor. “Our collaboration with ODS demonstrates how software-defined RAID can meet the rigorous demands of AI and HPC workloads, while optimizing costs and simplifying deployment.”
“With xiRAID at the heart of this solution, we have enabled the university to push its boundaries in research and innovation,” added Rakesh Sabharwal, founder of On Demand System. “The seamless integration of xiRAID with BeeGFS has ensured a reliable, high-performing storage backbone that matches the university’s ambitions.”
