# Is the GPU Half-Empty or Half-Full? Practical Scheduling Techniques for LLMs

arXiv: https://arxiv.org/abs/2410.17840 · HTML: https://mlsanity.com/paper/2410.17840/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 26 references that are in the map.
- First posted on arXiv 23 Oct 2024.

## Builds on (most cited first, 18 of 18)

- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [FlexGen: High-Throughput Generative Inference of Large Language Models with a Single GPU](https://mlsanity.com/paper/2303.06865/) (2023)
- [Splitwise: Efficient generative LLM inference using phase splitting](https://mlsanity.com/paper/2311.18677/) (2023)
- [Ray: A Distributed Framework for Emerging AI Applications](https://mlsanity.com/paper/1712.05889/) (2017)
- [Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve](https://mlsanity.com/paper/2403.02310/) (2024)
- [Clipper: A Low-Latency Online Prediction Serving System](https://mlsanity.com/paper/1612.03079/) (2016)
- [Llumnix: Dynamic Scheduling for Large Language Model Serving](https://mlsanity.com/paper/2406.03243/) (2024)
- [S-LoRA: Serving Thousands of Concurrent LoRA Adapters](https://mlsanity.com/paper/2311.03285/) (2023)
- [Learning Scheduling Algorithms for Data Processing Clusters](https://mlsanity.com/paper/1810.01963/) (2018)
- [Response Length Perception and Sequence Scheduling: An LLM-Empowered LLM Inference Pipeline](https://mlsanity.com/paper/2305.13144/) (2023)
- [S3: Increasing GPU Utilization during Generative Inference for Higher Throughput](https://mlsanity.com/paper/2306.06000/) (2023)
- [Fairness in Serving Large Language Models](https://mlsanity.com/paper/2401.00588/) (2023)
- [Serverless in the Wild: Characterizing and Optimizing the Serverless Workload at a Large Cloud Provider](https://mlsanity.com/paper/2003.03423/) (2020)
- [Extract-Transform-Load for Video Streams](https://mlsanity.com/paper/2310.04830/) (2023)
- [Optimizing Mixture of Experts using Dynamic Recompilations](https://mlsanity.com/paper/2205.01848/) (2022)
- [Stage: Query Execution Time Prediction in Amazon Redshift](https://mlsanity.com/paper/2403.02286/) (2024)
- [Blueprinting the Cloud: Unifying and Automatically Optimizing Cloud Data Infrastructures with BRAD -- Extended Version](https://mlsanity.com/paper/2407.15363/) (2024)

## Built on by (most cited first, 4 of 4)

- [Taming the Titans: A Survey of Efficient LLM Inference Serving](https://mlsanity.com/paper/2504.19720/) (2025)
- [TokenFlow: Responsive LLM Text Streaming Serving under Request Burst via Preemptive Scheduling](https://mlsanity.com/paper/2510.02758/) (2025)
- [OServe: Accelerating LLM Serving via Spatial-Temporal Workload Orchestration](https://mlsanity.com/paper/2602.12151/) (2026)
- [Auto-Scaling Heterogeneous Neural Processing Units for Energy and Cost-Efficient LLM Serving](https://mlsanity.com/paper/2607.16488/) (2026)

## Abstract (arXiv)

Serving systems for Large Language Models (LLMs) improve throughput by processing several requests concurrently. However, multiplexing hardware resources between concurrent requests involves non-trivial scheduling decisions. Practical serving systems typically implement these decisions at two levels: First, a load balancer routes requests to different servers which each hold a replica of the LLM. Then, on each server, an engine-level scheduler decides when to run a request, or when to queue or preempt it. Improved scheduling policies may benefit a wide range of LLM deployments and can often be implemented as "drop-in replacements" to a system's current policy. In this work, we survey scheduling techniques from the literature and from practical serving systems. We find that schedulers from the literature often achieve good performance but introduce significant complexity. In contrast, schedulers in practical deployments often leave easy performance gains on the table but are easy to implement, deploy and configure. This finding motivates us to introduce two new scheduling techniques, which are both easy to implement, and outperform current techniques on production workload traces.