# Echo: Error-Compensated 4-Bit Tensor-Parallel Communication for LLM Training and Inference

arXiv: https://arxiv.org/abs/2610.14277 · HTML: https://mlsanity.com/paper/2610.14277/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 51 of its 65 references that are in the map.
- First posted on arXiv 9 Oct 2026.

## Builds on (most cited first, 30 of 51)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [HellaSwag: Can a Machine Really Finish Your Sentence?](https://mlsanity.com/paper/1905.07830/) (2019)
- [FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness](https://mlsanity.com/paper/2205.14135/) (2022)
- [Pointer Sentinel Mixture Models](https://mlsanity.com/paper/1609.07843/) (2016)
- [PIQA: Reasoning about Physical Commonsense in Natural Language](https://mlsanity.com/paper/1911.11641/) (2019)
- [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://mlsanity.com/paper/2311.12022/) (2023)
- [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://mlsanity.com/paper/1909.08053/) (2019)
- [MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark](https://mlsanity.com/paper/2406.01574/) (2024)
- [AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration](https://mlsanity.com/paper/2306.00978/) (2023)
- [On Layer Normalization in the Transformer Architecture](https://mlsanity.com/paper/2002.04745/) (2020)
- [KIVI: A Tuning-Free Asymmetric 2bit Quantization for KV Cache](https://mlsanity.com/paper/2402.02750/) (2024)
- [KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantization](https://mlsanity.com/paper/2401.18079/) (2024)
- [Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research](https://mlsanity.com/paper/2402.00159/) (2024)
- [Data-Free Quantization Through Weight Equalization and Bias Correction](https://mlsanity.com/paper/1906.04721/) (2019)
- [Error Feedback Fixes SignSGD and other Gradient Compression Schemes](https://mlsanity.com/paper/1901.09847/) (2019)
- [PowerSGD: Practical Low-Rank Gradient Compression for Distributed Optimization](https://mlsanity.com/paper/1905.13727/) (2019)
- [FP8 Formats for Deep Learning](https://mlsanity.com/paper/2209.05433/) (2022)
- [BitNet: Scaling 1-bit Transformers for Large Language Models](https://mlsanity.com/paper/2310.11453/) (2023)
- [Adan: Adaptive Nesterov Momentum Algorithm for Faster Optimizing Deep Models](https://mlsanity.com/paper/2208.06677/) (2022)
- [EF21: A New, Simpler, Theoretically Better, and Practically Faster Error Feedback](https://mlsanity.com/paper/2106.05203/) (2021)
- [AI and Memory Wall](https://mlsanity.com/paper/2403.14123/) (2024)
- [Pretraining Large Language Models with NVFP4](https://mlsanity.com/paper/2509.25149/) (2025)
- [1-bit Adam: Communication Efficient Large-Scale Training with Adam's Convergence Speed](https://mlsanity.com/paper/2102.02888/) (2021)
- [Scaling Distributed Machine Learning with In-Network Aggregation](https://mlsanity.com/paper/1903.06701/) (2019)
- [The Curse of Depth in Large Language Models](https://mlsanity.com/paper/2502.05795/) (2025)

## Abstract (arXiv)

Tensor parallelism ends every layer with an all-reduce that sits between two dependent matrix multiplications, so its cost cannot be hidden and grows as compute outpaces interconnect bandwidth. Sending 4-bit instead of 16-bit values would cut this cost by nearly 4x, but the all-reduce must add the values it carries, and rounding the sum injects an error at every call. Model quality depends on how these errors accumulate over layers, yet existing methods treat each call on its own. We present Echo, a 4-bit all-reduce for training and inference that controls the accumulated error. Echo moves data in 4 bits and sums in high precision, so each value is quantized at most twice per call for any number of GPUs. The summation reveals each injected error, and Echo feeds it back into later calls, so old errors cancel instead of accumulating. We prove that the accumulated error stays bounded independently of depth. On A100 GPUs, Echo is 2.09x faster than the 16-bit NCCL all-reduce on 1 GB messages. It speeds up the time to first token by up to 1.16x and the training step by up to 1.09x on an idle interconnect, and by up to 1.44x and 1.38x as background traffic grows. It keeps the average accuracy over seven benchmarks within 0.4 points of 16-bit inference on Qwen3-8B. In 10B-token pretraining, its final loss is less than 0.9% above that of 16-bit training. Echo needs no FP8 or FP4 hardware.