Kinetic Campaign: Speeding Up Scientific Data Analytics with Computational Storage

Explore a 48-minute technical presentation that delves into innovative approaches for accelerating scientific data analytics using computational storage technologies. Learn about C2, a collaborative research project between Seagate and Los Alamos National Lab (LANL) designed for next-generation campaign storage. Discover how the integration of Kinetic Drives enables direct data analytics at the storage layer through asynchronous indexing, addressing challenges in large-scale data analytics and machine learning applications. Understand the complexities of managing massive datasets across tiered storage systems, including warm tiers with flash storage and cool tiers with high-density rotational drives. Gain insights into how asynchronously constructed indexes covering all data columns can significantly reduce query response times and data transfer requirements. Examine the unique challenges and solutions in combining computational storage technologies with erasure coding-based data protection schemes, particularly in environments where individual drives may have limited access to complete data records. Master the fundamentals of HPC storage architecture, explore use cases for large-scale scientific data analytics, and learn about emerging designs for computational storage in long-term archival data management.