Meta Andromeda Hits 1B Ads/Second in Retrieval, a 100x Lift Over 2024
jonathanmendez · x · 2026-11-05
Meta's engineering blog details the next generation of Andromeda, its personalized ads retrieval engine, built through co-design of model, system, and GPU hardware to bring near-ranking-level model quality to retrieval scale.
- Throughput: In production the system scores every eligible ad and surfaces top candidates at over one billion ads per second, a 100x lift over the 2024 Andromeda system and a new throughput benchmark for retrieval-stage models.
- Business impact: Since 2025, Andromeda has delivered a 2.4% increase in conversions on Facebook; advertisers get better matching across a larger candidate pool within the same latency budget.
- Architecture: Retrieval is the first stage of the ads recommendation pipeline, narrowing a vast pool to a small candidate set in milliseconds before a costlier ranking stage scores survivors in detail. Ranking-grade deep modeling was historically too expensive at retrieval scale.
- Approach: Building on the 2024 deployment on NVIDIA GH200 Grace Hopper Superchips, the team co-designed model, system, and hardware to push the efficiency frontier.
Authored by Haiyu Lu, Shuo Gu, and colleagues on Meta's engineering team.
More from Infra
- LFM3-Omni-2B ported to run fully offline on iPhone at ~80 tok/s with custom WebGPU kernels — JosephJacks_ · 2026-11-05
- Orbital Data Centers Launch to Sun-Synchronous Orbit, Memed as Elves Sailing West — rand_longevity · 2026-11-05
- Vietnam's Lâm Đồng to bring Starlink to mountain areas where poor signal hampers rescue — elonmusk · 2026-11-05
- SpaceX partners with Qualcomm for nationwide Starlink Mobile 5G network from 2027 — Polymarket · 2026-11-05
- Stopping runaway agent spend before the bill: pre-call guardrails and three gaps found the hard way — MaverikSh · 2026-11-05
- US AI scale faces two physical gates: delivered power and manufacturing — ingliguori · 2026-11-05