Earn Your CS Degree, Tuition-Free, 100% Online!
Learn Excel & Financial Modeling the Way Finance Teams Actually Use Them
Overview
Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off
One annual plan covers every course and certificate on Coursera. 40% off for a limited time.
Get Full Access
Explore a comprehensive demonstration of Keysight's AI fabric test methodology in this 29-minute technical presentation where Alex Bortek, Lead Product Manager at Keysight Technologies, showcases how to optimize AI infrastructure performance. Learn about the Kai Data Center Builder product that guides users through designing and building efficient AI fabrics, with emphasis on topology selection, collective operation algorithms, performance isolation, load balancing, and congestion control. Understand essential terminology including collective operations (broadcast, all-reduce, all-to-all), ranks, and data size metrics that determine performance. Witness a practical demonstration using a testbed of four 800-Gbps switches emulating 16 GPUs at 400 Gbps, which clearly illustrates how proper congestion control configuration can dramatically improve bandwidth utilization. Discover how to measure performance through metrics like collective completion time, algorithm bandwidth, and bus bandwidth to identify limiting factors in AI infrastructure. The presentation concludes with information about Keysight's Ultra Ethernet consortium membership and upcoming innovations in AI infrastructure testing. Recorded live in Santa Clara, California on April 25, 2025 as part of AI Infrastructure Field Day.
Syllabus
Demonstrating Keysight’s AI Fabric Test Methodology
Taught by
Tech Field Day