Meta shares Intelligent Delivery ads compute allocation system
Start here
Meta published details of Intelligent Delivery, an ML system that allocates compute across its ads serving stack toward a system-wide optimum. Meta says the system contributed roughly 10% incremental ranking performance on top of its model and system advancements from 2017 through the first half of 2026. At its core is a Resource Core Model with 6B sparse and 20M dense parameters that predicts each request's marginal utility; Meta says offline benchmarks showed about a 20% improvement in identifying high-impact ad requests. Multi-layered allocation mechanisms move resources across the system, and agentic AI increasingly adapts to shifts in models and hardware. Meta says agents contributed about 20% of Intelligent Delivery optimizations in the first half of 2026, and about 80% of the engineering effort on its most automated mechanism, from forming hypotheses to running experiments. Engineers still review and approve the final change before it ships.
Generated from reports · updated 12 h ago
Report timeline
Follow the reports to see the event from different sides.
- Meta Engineering BlogMeta Intelligent Delivery: ML System Allocating Compute for Ads Ranking
Meta detailed Intelligent Delivery, an ML-driven system that allocates compute across its ads serving stack toward the system-wide optimum, contributing roughly 10% incremental ranking performance on top of model and system advancements from 2017 through the first half of 2026. Its Resource Core Model uses 6B sparse and 20M dense parameters to predict each request's marginal utility, and Meta says it improved identification of high-impact ad requests by about 20% in offline benchmarks. The system also relies on multi-layered allocation mechanisms and increasingly agentic AI to adapt as models and hardware change.
Heat for this event
Not enough continuous observations yet to draw a trend.