Meta upgraded its ad service latency management by utilizing the open-source sched_ext framework. This change resulted in a 28% reduction in retrieval latency, improved ad ranking, and reduced energy consumption.
Meta's ads serving system manages over 5 million requests per second, totaling more than 400 billion daily. Latency is critical; reducing it enhances ad relevance and ROI for advertisers.
Meta previously employed general-purpose Linux kernel schedulers, like CFS and EEVDF, which did not consider specific workload characteristics. This generic approach led to increased latency issues, particularly with the kernel upgrade to v6.9.
To combat the latency problem, Meta partnered with sched_ext's developers to create a customized scheduling policy tailored for ads delivery. This BPF-based framework allows optimization by prioritizing latency-sensitive tasks specific to ads.
The implementation of sched_ext resulted in a decrease in the 99th percentile latency by 28%, alongside a 3.28 MW power reduction. Moreover, there was a 1.1% increase in the number of ads ranked, demonstrating the effectiveness of workload-specific optimizations.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's briefNew every morning, and the back catalogue is archived by date.
Meta upgraded its ad service latency management by utilizing the open-source sched_ext framework. This change resulted in a 28% reduction in retrieval latency, improved ad ranking, and reduced energy consumption.