Christopher Zimmer

The Lustre filesystem serves as a vital element in high-performance parallel storage, meeting the rising demands of scientific, research, and enterprise environments. Widely deployed across HPC environments, ranging from small-scale applications in AI/ML, to domains like oil and gas, drug discovery, and meteorology, and manufacturing, Lustre addresses the universal challenge of efficiently accessing vast and ever-increasing volumes of data. Lustre is the filesystem of choice on six out of the top 10 fastest supercomputers in the world today, over 65% of the top 100, and also for over 60% of the top 500. Despite its widespread popularity, there is a lack of a complete and up-to-date reference, covering Lustre’s evolution, design, and various advancements made over the years. In this journal, we aim to fill this gap by providing a comprehensive journey of Lustre, including its history with significant contributions to HPC, detailed architecture and design elements, exploration of advancements added through its evolution, and future directions. Additionally, we present a comparison of Lustre with other prominent storage technologies of the era. To illustrate the current state of Lustre, we analyze several filesystem trends, including utilization, performance, and usage patterns on Orion, the Lustre filesystem on the first exascale supercomputer Frontier. We hope that this journal serves as a comprehensive educational reference for the current and future generations interested in HPC filesystem storage aspects.

This report highlights significant strides made by Oak Ridge National Laboratory’s Oak Ridge LeadershipComputing Facility (OLCF) in advancing computational research and infrastructure. Through the AdvancedComputing Ecosystem (ACE) strategic initiative, OLCF has been successfully integrated with DOE’sIntegrated Research Infrastructure (IRI) program, establishing itself as a critical framework for enhancingscientific computing capabilities across various domains. This report outlines the activities, accomplishments,and future directions of ACE, emphasizing its role in developing cutting-edge technologies, supportingscience pilots, and fostering collaborations that drive scientific innovation.One of the key achievements of ACE in FY24 has been the establishment of a robust testbed environmentthat facilitates the development and validation of new computing technologies. This testbed, integratedinto the larger IRI ecosystem through Energy Sciences Network (ESnet), allows for seamless collaborationamong DOE’s Advanced Scientific Computing Research (ASCR) facilities, including the Argonne LeadershipComputing Facility (ALCF), the National Energy Research Scientific Computing Center (NERSC), andJefferson Lab (JLab). The testbed has been instrumental in enabling science pilots to experiment withinnovative methodologies in a controlled environment, thereby accelerating the translation of research intopractical applications.ACE has been instrumental in advancing a series of science pilots using computational resources fromOLCF. These pilots span a wide range of scientific domains, from basic energy sciences to biological andenvironmental research, demonstrating the versatility and impact of the ACE framework. By integratingcutting-edge technologies, such as AI-driven analytics and real-time data processing, with HPC, these pilotshave enabled significant advancements in scientific workflows. For instance, projects like LCLStreamand DELERIA have showcased the potential of ACE in optimizing experimental designs, acceleratingdata analysis, and driving innovations that are critical for addressing some of the most complex scientificchallenges of IRI pathfinder projects.ACE has also made significant progress in foundational technology development, particularly in areas suchas data movement, interface design, and scheduling. These advancements have been key in addressingthe complex requirements of modern scientific workflows, ensuring that OLCF remains at the forefront ofhigh-performance computing (HPC). The report details several ongoing projects, including the developmentof the OLCF Facility API, Data Streaming technologies, and the Zambeze distributed orchestration system,which together provide a secure and flexible framework for managing computational resources across adistributed infrastructure.In addition to technical achievements, ACE has played a central role in outreach and engagement activi-ties. Through participation in conferences, hackathons, and training sessions, the ACE team has activelycontributed to the broader scientific community. These efforts have not only facilitated the exchange ofknowledge but have also strengthened collaborations with other DOE facilities and research institutions,further enhancing the impact of the ACE program.Looking ahead, the report outlines strategic plans for FY25, which include continued support for sciencepilots, further integration with ASCR facilities, and the exploration of new technologies. These initiatives aredesigned to ensure that ACE remains a dynamic and adaptable platform, capable of meeting the evolvingneeds of the scientific community. The ongoing collaboration with ORNL’s INTERSECT initiative and theplanned technology refreshes underscore ACE’s commitment to maintaining its leadership in HPC.

Modern high-performance computing systems have multiple GPUs and network interface cards (NICs) per node. The resulting network architectures have multilevel hierarchies of subnetworks with different interconnect and software technologies. These systems offer multiple vendor-provided communication capabilities and library implementations (IPC, MPI, NCCL, RCCL, OneCCL) with APIs providing varying levels of performance across the different levels. Understanding this performance is currently difficult because of the wide range of architectures and programming models (CUDA, HIP, OneAPI). We present CommBench, a library with cross-system portability and a high-level API that enables developers to easily build microbenchmarks relevant to their use cases and gain insight into the performance (bandwidth & latency) of multiple implementation libraries on different networks. We demonstrate CommBench with three sets of microbenchmarks that profile the performance of six systems. Our experimental results reveal the effect of multiple NICs on optimizing the bandwidth across nodes and also present the performance characteristics of four available communication libraries within and across nodes of NVIDIA, AMD, and Intel GPU networks.