Higher education and research organizations require a processing and storage infrastructure capable of running high performance computing (HPC) and artificial intelligence (AI) workloads and efficiently storing its data. Traditional file systems have bottlenecks when reading and writing large amounts of data, especially those of I/O intensive workloads that require concurrent processing with multi-client nodes communication.
Quanta Cloud Technology (QCT) is a leading cloud data center and HPC solution provider, dedicated to providing purpose-built solutions with best-fit compute, storage, and networking building blocks as a solution to meet I/O intensive workloads. This paper describes how QCT partnered with ThinkParQ to develop its QCT BeeGFS software-defined parallel-cluster file system that avoids architectural bottlenecks. The QCT storage building block powered by 3rd Generation Intel® Xeon® Scalable processors features Intel® Software Guard Extensions (SGX), Intel® Crypto Acceleration, Intel® Total Memory Encryption (TME), and Intel Platform Firmware Resilience (PFR) to enable enhanced collaboration using shared data without compromising privacy while also meeting HPC and AI storage requirements.