Exploring Hierarchical Interest Representation For Meta Ads Deep Funnel Optimization

Engineering4d ago

Modernizing the Meta Ads Service With an Open-Source Kernel Scheduler

To address latency regressions and optimize ads retrieval performance, Meta's Ads and Linux Kernel teams collaborated to implement a customized scheduling policy using the open-source, BPF-based extensible scheduling framework, sched_ext. This resulted in a 28% reduction in ads retrieval stage tail latency and 3.28 MW power savings. By utilizing sched_ext, Meta was able to decouple scheduler optimizations from upstream Linux kernel releases, enabling continuous improvement and reduced experimentation costs.

SocialScale

1 min

Exploring Hierarchical Interest Representation For Meta Ads Deep Funnel Optimization

Modernizing the Meta Ads Service With an Open-Source Kernel Scheduler

Meta&#8217;s AI Storage Blueprint at Scale

Meta’s AI Storage Blueprint at Scale