Optimizing Storage for Meta's Trillion-Video Catalog: Achieving Pareto Efficiency

on YouTube ↗16m

" summary

Covers Meta’s strategy for reducing storage footprint across a trillion+ video catalog that grows by over a billion videos per day, focusing on reaching Pareto-efficient tradeoffs between space and delivered video quality. Describes the core technical approach and practical strategies used to improve storage efficiency without degrading quality, and summarizes lessons learned and ongoing system refinements for scalable video infrastructure.

AI-written summary · about

" related talks

• 33mArchitecting Scalable Multi-Modal Video SearchSebastiano Milardo, Pietro Piccini · PyData Paris 2025
• 15mAn Efficient Cloud Storage Model with Compacted Metadata Management for Performance...Kai Zhang, Zili Shao, Tianyu Wang · FAST '26
• 12mRevolutionizing Data ManagementVenkatachala Nivas Chainuru · Conf42 Machine Learning 2025
• 01h 19mThe Future of the Past: Challenges in Archival StorageEthan L. Miller · USENIX ATC '20
• 14mRe-think Data Management Software Design Upon the Arrival of Storage Hardware withTong Zhang, Yang Liu, Qi Wu, Xubin Chen, Ning Zheng, Jiangpeng Li, Yong Peng, Fei Sun, Hao Zhong · 12th USENIX Workshop on Hot Topics in Storage and File Systems (HotStorage '20)
• 47mMerkle Forest, or How We Reduced Metadata Size and Also Speeded Up Duplicate SearchSergey Li · HighLoad++ 2024
• 32mThe Future of Metadata in Prometheus: Enhancing Storage and UsabilityGanesh Vernekar, Arve Knudsen · PromCon EU 2024
• 16mTVStore: Automatically Bounding Time Series Storage via Time-Varying CompressionJianmin Wang, Yanzhe An, Yue Su, Yuqing Zhu · FAST '22