# Dovetail: A CPU/GPU Heterogeneous Speculative Decoding for LLM inference

arXiv: https://arxiv.org/abs/2412.18934 · HTML: https://mlsanity.com/paper/2412.18934/

## Key facts

- Cited by 13 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 28 of its 44 references that are in the map.
- First posted on arXiv 25 Dec 2024.

## Placed in

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

## Builds on (most cited first, 28 of 28)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers](https://mlsanity.com/paper/2210.17323/) (2022)
- [Fast Inference from Transformers via Speculative Decoding](https://mlsanity.com/paper/2211.17192/) (2022)
- [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)
- [KIVI: A Tuning-Free Asymmetric 2bit Quantization for KV Cache](https://mlsanity.com/paper/2402.02750/) (2024)
- [TinyLlama: An Open-Source Small Language Model](https://mlsanity.com/paper/2401.02385/) (2024)
- [EAGLE: Speculative Sampling Requires Rethinking Feature Uncertainty](https://mlsanity.com/paper/2401.15077/) (2024)
- [The State of Sparsity in Deep Neural Networks](https://mlsanity.com/paper/1902.09574/) (2019)
- [Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activations](https://mlsanity.com/paper/1609.07061/) (2016)
- [Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning](https://mlsanity.com/paper/2310.06694/) (2023)
- [EAGLE-2: Faster Inference of Language Models with Dynamic Draft Trees](https://mlsanity.com/paper/2406.16858/) (2024)
- [Unlocking Efficiency in Large Language Model Inference: A Comprehensive Survey of Speculative Decoding](https://mlsanity.com/paper/2401.07851/) (2024)
- [Gated Multimodal Units for Information Fusion](https://mlsanity.com/paper/1702.01992/) (2017)
- [PowerInfer: Fast Large Language Model Serving with a Consumer-grade GPU](https://mlsanity.com/paper/2312.12456/) (2023)
- [Speculative Decoding: Exploiting Speculative Execution for Accelerating Seq2seq Generation](https://mlsanity.com/paper/2203.16487/) (2022)
- [Speculative Decoding with Big Little Decoder](https://mlsanity.com/paper/2302.07863/) (2023)
- [Accelerating LLM Inference with Staged Speculative Decoding](https://mlsanity.com/paper/2308.04623/) (2023)
- [Insertion Transformer: Flexible Sequence Generation via Insertion Operations](https://mlsanity.com/paper/1902.03249/) (2019)
- [f-Divergence Minimization for Sequence-Level Knowledge Distillation](https://mlsanity.com/paper/2307.15190/) (2023)
- [MagicDec: Breaking the Latency-Throughput Tradeoff for Long Context Generation with Speculative Decoding](https://mlsanity.com/paper/2408.11049/) (2024)
- [Cascade Speculative Drafting for Even Faster LLM Inference](https://mlsanity.com/paper/2312.11462/) (2023)
- [SpecExec: Massively Parallel Speculative Decoding for Interactive LLM Inference on Consumer Devices](https://mlsanity.com/paper/2406.02532/) (2024)
- [KV Cache Compression, But What Must We Give in Return? A Comprehensive Benchmark of Long Context Capable Approaches](https://mlsanity.com/paper/2407.01527/) (2024)
- [ENGINE: Energy-Based Inference Networks for Non-Autoregressive Machine Translation](https://mlsanity.com/paper/2005.00850/) (2020)
- [ConsistentEE: A Consistent and Hardness-Guided Early Exiting Method for Accelerating Language Models Inference](https://mlsanity.com/paper/2312.11882/) (2023)

## Built on by (most cited first, 13 of 13)

- [Speculative Decoding and Beyond: An In-Depth Survey of Techniques](https://mlsanity.com/paper/2502.19732/) (2025)
- [DREAM: Drafting with Refined Target Features and Entropy-Adaptive Cross-Attention Fusion for Multimodal Speculative Decoding](https://mlsanity.com/paper/2505.19201/) (2025)
- [LittleBit: Ultra Low-Bit Quantization via Latent Factorization](https://mlsanity.com/paper/2506.13771/) (2025)
- [TailorKV: A Hybrid Framework for Long-Context Inference via Tailored KV Cache Optimization](https://mlsanity.com/paper/2505.19586/) (2025)
- [LayerScope: Predictive Cross-Layer Scheduling for Efficient Multi-Batch MoE Inference on Legacy Servers](https://mlsanity.com/paper/2509.23638/) (2025)
- [A Survey on Collaborating Small and Large Language Models for Performance, Cost-effectiveness, Cloud-edge Privacy, and Trustworthiness](https://mlsanity.com/paper/2510.13890/) (2025)
- [Nightjar: Dynamic Adaptive Speculative Decoding for Large Language Models Serving](https://mlsanity.com/paper/2512.22420/) (2025)
- [Sparrow: Text-Anchored Window Attention with Visual-Semantic Glimpsing for Speculative Decoding in Video LLMs](https://mlsanity.com/paper/2602.15318/) (2026)
- [Speculate Deep and Accurate: Lossless and Training-Free Acceleration for Offloaded LLMs via Substitute Speculative Decoding](https://mlsanity.com/paper/2509.18344/) (2025)
- [Rhea: Role-aware Heuristic Episodic Attention for Conversational LLMs](https://mlsanity.com/paper/2512.06869/) (2025)
- [Compiler-Assisted Speculative Sampling for Accelerated LLM Inference on Heterogeneous Edge Devices](https://mlsanity.com/paper/2602.08060/) (2026)
- [WANSpec: Leveraging Global Compute Capacity for LLM Inference](https://mlsanity.com/paper/2602.18931/) (2026)
- [Role-aware Heuristic Episodic Attention for Conversational LLMs](https://mlsanity.com/paper/2610.00958/) (2026)

## Abstract (arXiv)

With the continuous advancement in the performance of large language models (LLMs), their demand for computational resources and memory has significantly increased, which poses major challenges for efficient inference on consumer-grade devices and legacy servers. These devices typically feature relatively weaker GPUs and stronger CPUs. Although techniques such as parameter offloading and partial offloading can alleviate GPU memory pressure to some extent, their effectiveness is limited due to communication latency and suboptimal hardware resource utilization. To address this issue, we propose Dovetail, a lossless inference acceleration method that leverages the complementary characteristics of heterogeneous devices and the advantages of speculative decoding. Dovetail deploys a draft model on the GPU to perform preliminary predictions, while a target model running on the CPU validates these outputs. By reducing the granularity of data transfer, Dovetail significantly minimizes communication overhead. To further improve efficiency, we optimize the draft model specifically for heterogeneous hardware environments by reducing the number of draft tokens to lower parallel verification latency, increasing model depth to enhance predictive capabilities, and introducing a Dynamic Gating Fusion (DGF) mechanism to improve the integration of feature and embedding information. We conduct comprehensive evaluations of Dovetail across various consumer-grade GPUs, covering multiple tasks and mainstream models. Experimental results on 13B models demonstrate that Dovetail achieves inference speedups ranging from 1.79x to 10.1x across different devices, while maintaining consistency and stability in the distribution of generated texts.