# DCAF: A Dynamic Computation Allocation Framework for Online Serving System

arXiv: https://arxiv.org/abs/2006.09684 · HTML: https://mlsanity.com/paper/2006.09684/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Jun 2020.

## Built on by (most cited first, 7 of 7)

- [Adaptive Neural Ranking Framework: Toward Maximized Business Goal for Cascade Ranking Systems](https://mlsanity.com/paper/2310.10462/) (2023)
- [Computation Resource Allocation Solution in Recommender Systems](https://mlsanity.com/paper/2103.02259/) (2021)
- [Improve ROI with Causal Learning and Conformal Prediction](https://mlsanity.com/paper/2407.01065/) (2024)
- [RPAF: A Reinforcement Prediction-Allocation Framework for Cache Allocation in Large-Scale Recommender Systems](https://mlsanity.com/paper/2409.13175/) (2024)
- [GreenFlow: A Computation Allocation Framework for Building Environmentally Sound Recommendation System](https://mlsanity.com/paper/2312.16176/) (2023)
- [Adaptive2: Adaptive Domain Mining for Fine-grained Domain Adaptation Modeling](https://mlsanity.com/paper/2412.08198/) (2024)
- [MaRCA: Multi-Agent Reinforcement Learning for Dynamic Computation Allocation in Large-Scale Recommender Systems](https://mlsanity.com/paper/2512.24325/) (2025)

## Abstract (arXiv)

Modern large-scale systems such as recommender system and online advertising system are built upon computation-intensive infrastructure. The typical objective in these applications is to maximize the total revenue, e.g. GMV~(Gross Merchandise Volume), under a limited computation resource. Usually, the online serving system follows a multi-stage cascade architecture, which consists of several stages including retrieval, pre-ranking, ranking, etc. These stages usually allocate resource manually with specific computing power budgets, which requires the serving configuration to adapt accordingly. As a result, the existing system easily falls into suboptimal solutions with respect to maximizing the total revenue. The limitation is due to the face that, although the value of traffic requests vary greatly, online serving system still spends equal computing power among them. In this paper, we introduce a novel idea that online serving system could treat each traffic request differently and allocate "personalized" computation resource based on its value. We formulate this resource allocation problem as a knapsack problem and propose a Dynamic Computation Allocation Framework~(DCAF). Under some general assumptions, DCAF can theoretically guarantee that the system can maximize the total revenue within given computation budget. DCAF brings significant improvement and has been deployed in the display advertising system of Taobao for serving the main traffic. With DCAF, we are able to maintain the same business performance with 20% computation resource reduction.