# SwiftSpec: Ultra-Low Latency LLM Decoding by Scaling Asynchronous Speculative Decoding

arXiv: https://arxiv.org/abs/2506.11309 · HTML: https://mlsanity.com/paper/2506.11309/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Placed in Speculative decoding, one of 12 directions in Efficient LLM inference and serving.
- Builds on 29 of its 39 references that are in the map.
- First posted on arXiv 12 Jun 2025.

## Placed in

- Efficient LLM inference and serving / [Speculative decoding](https://mlsanity.com/inference/speculative-decoding/)

## Builds on (most cited first, 29 of 29)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://mlsanity.com/paper/1909.08053/) (2019)
- [Fast Inference from Transformers via Speculative Decoding](https://mlsanity.com/paper/2211.17192/) (2022)
- [SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models](https://mlsanity.com/paper/2211.10438/) (2022)
- [DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence](https://mlsanity.com/paper/2401.14196/) (2024)
- [Accelerating Large Language Model Decoding with Speculative Sampling](https://mlsanity.com/paper/2302.01318/) (2023)
- [Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Heads](https://mlsanity.com/paper/2401.10774/) (2024)
- [Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyond](https://mlsanity.com/paper/1602.06023/) (2016)
- [DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving](https://mlsanity.com/paper/2401.09670/) (2024)
- [Splitwise: Efficient generative LLM inference using phase splitting](https://mlsanity.com/paper/2311.18677/) (2023)
- [Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve](https://mlsanity.com/paper/2403.02310/) (2024)
- [EAGLE-2: Faster Inference of Language Models with Dynamic Draft Trees](https://mlsanity.com/paper/2406.16858/) (2024)
- [SpecInfer: Accelerating Generative Large Language Model Serving with Tree-based Speculative Inference and Verification](https://mlsanity.com/paper/2305.09781/) (2023)
- [CacheGen: KV Cache Compression and Streaming for Fast Large Language Model Serving](https://mlsanity.com/paper/2310.07240/) (2023)
- [LongCoder: A Long-Range Pre-trained Language Model for Code Completion](https://mlsanity.com/paper/2306.14893/) (2023)
- [DAPPLE: A Pipelined Data Parallel Approach for Training Large Models](https://mlsanity.com/paper/2007.01045/) (2020)
- [PEARL: Parallel Speculative Decoding with Adaptive Draft Length](https://mlsanity.com/paper/2408.11850/) (2024)
- [Code-as-Monitor: Constraint-aware Visual Programming for Reactive and Proactive Robotic Failure Detection](https://mlsanity.com/paper/2412.04455/) (2024)
- [When Search Engine Services meet Large Language Models: Visions and Challenges](https://mlsanity.com/paper/2407.00128/) (2024)
- [Language Models for Code Completion: A Practical Evaluation](https://mlsanity.com/paper/2402.16197/) (2024)
- [TurboSpec: Closed-loop Speculation Control System for Optimizing LLM Serving Goodput](https://mlsanity.com/paper/2406.14066/) (2024)
- [Large Language Models vs. Search Engines: Evaluating User Preferences Across Varied Information Retrieval Scenarios](https://mlsanity.com/paper/2401.05761/) (2024)
- [PipeInfer: Accelerating LLM Inference using Asynchronous Pipelined Speculation](https://mlsanity.com/paper/2407.11798/) (2024)
- [AMUSD: Asynchronous Multi-Device Speculative Decoding for LLM Acceleration](https://mlsanity.com/paper/2410.17375/) (2024)
- [Prepacking: A Simple Method for Fast Prefilling and Increased Throughput in Large Language Models](https://mlsanity.com/paper/2404.09529/) (2024)

## Built on by (most cited first, 9 of 9)

- [Speculative Speculative Decoding](https://mlsanity.com/paper/2603.03251/) (2026)
- [MoE-SpeQ: Speculative Quantized Decoding with Proactive Expert Prefetching and Offloading for Mixture-of-Experts](https://mlsanity.com/paper/2511.14102/) (2025)
- [Fast LLM Post-training via Decoupled and Fastest-of-N Speculation](https://mlsanity.com/paper/2511.16193/) (2025)
- [Make Every Draft Count: Hidden State based Speculative Decoding](https://mlsanity.com/paper/2602.21224/) (2026)
- [A Pipelined Collaborative Speculative Decoding Framework for Efficient Edge-Cloud LLM Inference](https://mlsanity.com/paper/2603.19133/) (2026)
- [Unlocking Parallelism in Autoregressive Language Models via Speculative Decoding with Progressive Tree Drafting](https://mlsanity.com/paper/2607.10661/) (2026)
- [HySparse2: Hybrid Sparse Attention with Two-Level KV Sharing](https://mlsanity.com/paper/2609.26368/) (2026)
- [ODIA: Oriented Distillation for Inline Acceleration of LLM-based Function Calling](https://mlsanity.com/paper/2507.08877/) (2025)
- [PRISM: Parametrically Refactoring Inference for Speculative Sampling Draft Models](https://mlsanity.com/paper/2602.01762/) (2026)

## Abstract (arXiv)

Low-latency decoding for large language models (LLMs) is crucial for applications like chatbots and code assistants, yet generating long outputs remains slow in single-query settings. Prior work on speculative decoding (which combines a small draft model with a larger target model) and tensor parallelism has each accelerated decoding. However, conventional approaches fail to apply both simultaneously due to imbalanced compute requirements (between draft and target models), KV-cache inconsistencies, and communication overheads under small-batch tensor-parallelism. This paper introduces SwiftSpec, a system that targets ultra-low latency for LLM decoding. SwiftSpec redesigns the speculative decoding pipeline in an asynchronous and disaggregated manner, so that each component can be scaled flexibly and remove draft overhead from the critical path. To realize this design, SwiftSpec proposes parallel tree generation, tree-aware KV cache management, and fused, latency-optimized kernels to overcome the challenges listed above. Across 5 model families and 6 datasets, SwiftSpec achieves an average of 1.75x speedup over state-of-the-art speculative decoding systems and, as a highlight, serves Llama3-70B at 348 tokens/s on 8 Nvidia Hopper GPUs, making it the fastest known system for low-latency LLM serving at this scale.