# Seesaw: High-throughput LLM Inference via Model Re-sharding

arXiv: https://arxiv.org/abs/2503.06433 · HTML: https://mlsanity.com/paper/2503.06433/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- Placed in Serving systems and request scheduling, one of 12 directions in Efficient LLM inference and serving.
- Builds on 44 of its 78 references that are in the map.
- First posted on arXiv 9 Mar 2025.

## Placed in

- Efficient LLM inference and serving / [Serving systems and request scheduling](https://mlsanity.com/inference/serving-systems-and-request-scheduling/)

## Builds on (most cited first, 30 of 44)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Code Llama: Open Foundation Models for Code](https://mlsanity.com/paper/2308.12950/) (2023)
- [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://mlsanity.com/paper/1909.08053/) (2019)
- [From Local to Global: A Graph RAG Approach to Query-Focused Summarization](https://mlsanity.com/paper/2404.16130/) (2024)
- [ZeRO: Memory Optimizations Toward Training Trillion Parameter Models](https://mlsanity.com/paper/1910.02054/) (2019)
- [PyTorch FSDP: Experiences on Scaling Fully Sharded Data Parallel](https://mlsanity.com/paper/2304.11277/) (2023)
- [DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving](https://mlsanity.com/paper/2401.09670/) (2024)
- [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)
- [Efficiently Scaling Transformer Inference](https://mlsanity.com/paper/2211.05102/) (2022)
- [Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve](https://mlsanity.com/paper/2403.02310/) (2024)
- [A Discourse-Aware Attention Model for Abstractive Summarization of Long Documents](https://mlsanity.com/paper/1804.05685/) (2018)
- [Ring Attention with Blockwise Transformers for Near-Infinite Context](https://mlsanity.com/paper/2310.01889/) (2023)
- [ZeRO-Offload: Democratizing Billion-Scale Model Training](https://mlsanity.com/paper/2101.06840/) (2021)
- [LongVILA: Scaling Long-Context Visual Language Models for Long Videos](https://mlsanity.com/paper/2408.10188/) (2024)
- [PowerInfer: Fast Large Language Model Serving with a Consumer-grade GPU](https://mlsanity.com/paper/2312.12456/) (2023)
- [SARATHI: Efficient LLM Inference by Piggybacking Decodes with Chunked Prefills](https://mlsanity.com/paper/2308.16369/) (2023)
- [Can Foundation Models Wrangle Your Data?](https://mlsanity.com/paper/2205.09911/) (2022)
- [Demystifying Parallel and Distributed Deep Learning: An In-Depth Concurrency Analysis](https://mlsanity.com/paper/1802.09941/) (2018)
- [LLM Inference Unveiled: Survey and Roofline Model Insights](https://mlsanity.com/paper/2402.16363/) (2024)
- [Sequence Parallelism: Long Sequence Training from System Perspective](https://mlsanity.com/paper/2105.13120/) (2021)
- [Inference without Interference: Disaggregate LLM Inference for Mixed Downstream Workloads](https://mlsanity.com/paper/2401.11181/) (2024)
- [DeepSpeed-FastGen: High-throughput Text Generation for LLMs via MII and DeepSpeed-Inference](https://mlsanity.com/paper/2401.08671/) (2024)
- [Towards Efficient Generative Large Language Model Serving: A Survey from Algorithms to Systems](https://mlsanity.com/paper/2312.15234/) (2023)
- [NanoFlow: Towards Optimal Large Language Model Serving Throughput](https://mlsanity.com/paper/2408.12757/) (2024)
- [Petals: Collaborative Inference and Fine-tuning of Large Models](https://mlsanity.com/paper/2209.01188/) (2022)

## Built on by (most cited first, 15 of 15)

- [Hardware-Efficient Attention for Fast Decoding](https://mlsanity.com/paper/2505.21487/) (2025)
- [ElasticMM: Efficient Multimodal LLMs Serving with Elastic Multimodal Parallelism](https://mlsanity.com/paper/2507.10069/) (2025)
- [FLYING SERVING: On-the-Fly Parallelism Switching for Large Language Model Serving](https://mlsanity.com/paper/2602.22593/) (2026)
- [Military AI Cyber Agents (MAICAs) Constitute a Global Threat to Critical Infrastructure](https://mlsanity.com/paper/2506.12094/) (2025)
- [Adaptive and Resource-efficient Agentic AI Systems for Mobile and Embedded Devices: A Survey](https://mlsanity.com/paper/2510.00078/) (2025)
- [10Cache: Heterogeneous Resource-Aware Tensor Caching and Migration for LLM Training](https://mlsanity.com/paper/2511.14124/) (2025)
- [Scepsy: Serving Agentic Workflows Using Aggregate LLM Pipelines](https://mlsanity.com/paper/2604.15186/) (2026)
- [Attention to Detail: Evaluating Energy, Performance, and Accuracy Trade-offs Across vLLM Configurations](https://mlsanity.com/paper/2607.09172/) (2026)
- [Learning to Shard: RL for Co-optimizing the Parallelism Degrees and Per-operator Sharding Dimensions in Distributed LLM Inference](https://mlsanity.com/paper/2509.00217/) (2025)
- [DCC: Data-Centric Compilation of Machine Learning Kernels for Processing-In-Memory Architectures](https://mlsanity.com/paper/2511.15503/) (2025)
- [PipeMax: Enhancing Offline LLM Inference on Commodity GPU Servers](https://mlsanity.com/paper/2605.02189/) (2026)
- [Beyond Tokens: A Survey on Decoding Methods for Large Language and Vision-Language Models](https://mlsanity.com/paper/2608.14797/) (2026)
- [KV-Pipe: On the Relation Between KV Sharing and Pipeline Parallel Efficiency in LLMs](https://mlsanity.com/paper/2608.15943/) (2026)
- [Efficient Vision-Language-Action Management and Serving for Robot Factories](https://mlsanity.com/paper/2609.12075/) (2026)
- [SPLASH: Switching Parallel Layouts of Attention with Seamless Handoff for LLM Serving](https://mlsanity.com/paper/2609.37626/) (2026)

## Abstract (arXiv)

To improve the efficiency of distributed large language model (LLM) inference, various parallelization strategies, such as tensor and pipeline parallelism, have been proposed. However, the distinct computational characteristics inherent in the two stages of LLM inference-prefilling and decoding-render a single static parallelization strategy insufficient for the effective optimization of both stages. In this work, we present Seesaw, an LLM inference engine optimized for throughput-oriented tasks. The key idea behind Seesaw is dynamic model re-sharding, a technique that facilitates the dynamic reconfiguration of parallelization strategies across stages, thereby maximizing throughput at both phases. To mitigate re-sharding overhead and optimize computational efficiency, we employ tiered KV cache buffering and transition-minimizing scheduling. These approaches work synergistically to reduce the overhead caused by frequent stage transitions while ensuring maximum batching efficiency. Our evaluation demonstrates that Seesaw achieves a throughput increase of up to 1.78x (1.36x on average) compared to vLLM, the most widely used state-of-the-art LLM inference engine.