John Bloss, HPC Deployment Manager at Lambda, led a comprehensive webinar on benchmarking and validation in system performance optimization, particularly for AI and ML systems, emphasizing the importance of both benchmarking for ideal performance expectations and validation for user-specific needs. The session covered various performance testing methods, including CPU and memory benchmarks, network testing, and GPU-specific diagnostics, while highlighting the critical role of efficient resource use in managing high-cost components. The conversation ended with a discussion of an upcoming competition focused on CPU performance testing, including rules and scheduling details for participating teams.
SUBMIT • ATTEND • VOLUNTEER
https://sc25.supercomputing.org/
—
SC25 • 16–21 NOV 2025 • ST. LOUIS, MO
The International Conference for High Performance Computing, Networking, Storage, and Analysis
