Watch a 20-minute conference talk from Open Compute Project exploring how hyperscale system designs can simultaneously achieve better efficiency and performance. Learn from Erik Riedel and Sean Magann as they analyze trade-offs between OCP hyperscale systems and traditional server/rack designs, demonstrating how reduced complexity leads to improved scalability and lower costs. Discover how innovative collaborative design approaches crossing hardware, software and operations have created solutions that minimize resource consumption while maximizing performance. Examine the environmental benefits of these design choices, including reduced carbon footprints in both operations and scope 3 emissions, particularly as systems remain in productive use over time. Understand how prioritizing simplicity and efficiency in data center infrastructure design generates compounding benefits through decreased energy usage, materials consumption, and operational overhead.
Scaling Down While Scaling Up - Design Choices That Increase Efficiency and Performance
Open Compute Project via YouTube
Overview
Syllabus
Scaling Down While Scaling Up – Design Choices That Increase Efficiency & Performance
Taught by
Open Compute Project