Performance @Scale 2019 recap
Aug 9, 2019Marty Greenia
Summary
This post summarizes presentations from the @Scale 2019 conference, highlighting advancements in performance engineering across various domains. Key contributions include methodologies for analyzing Facebook AI workloads on accelerated platforms (TPUs/GPUs), techniques for scaling ML models, and strategies for improving mobile app performance (app size, startup, crash rates). It also covers open-source browser contributions to enhance web app performance and the development of tools like FlameScope for detailed profiler analysis. Furthermore, it details efforts in monitoring real user perceived performance on native apps and scalable approaches to generating iOS binary order files for improved startup performance.
More from the Scalability and Large-Scale Systems Engineering trail
- Open-Sourcing Rebalancer: A Generic, High-Performance Library for Solving Assignment Problems
- ZGateway: Learnings from Putting a Proxy in Front of ZippyDB
- MTIA 300: Meta’s First Training Chip with Built-in NICs and Communication-Offloading Engines
- From User Sequences to Scaling Laws: A Multi-Stage Architecture for Meta’s Ads Ranking
- GEM Training: How Meta Doubled the Efficiency of Its LLM-Scale Ads Foundation Model