# MX+: Pushing the Limits of Microscaling Formats for Efficient Large Language Model Serving

arXiv: https://arxiv.org/abs/2510.14557 · HTML: https://mlsanity.com/paper/2510.14557/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Placed in KV cache quantization, one of 12 directions in Efficient LLM inference and serving.
- Builds on 32 of its 71 references that are in the map.
- First posted on arXiv 16 Oct 2025.

## Placed in

- Efficient LLM inference and serving / [KV cache quantization](https://mlsanity.com/inference/kv-cache-quantization/)

## Builds on (most cited first, 30 of 32)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [ImageNet Large Scale Visual Recognition Challenge](https://mlsanity.com/paper/1409.0575/) (2014)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [Training data-efficient image transformers & distillation through attention](https://mlsanity.com/paper/2012.12877/) (2020)
- [OPT: Open Pre-trained Transformer Language Models](https://mlsanity.com/paper/2205.01068/) (2022)
- [QLoRA: Efficient Finetuning of Quantized LLMs](https://mlsanity.com/paper/2305.14314/) (2023)
- [Pointer Sentinel Mixture Models](https://mlsanity.com/paper/1609.07843/) (2016)
- [SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models](https://mlsanity.com/paper/2211.10438/) (2022)
- [Phi-4 Technical Report](https://mlsanity.com/paper/2412.08905/) (2024)
- [QuaRot: Outlier-Free 4-Bit Inference in Rotated LLMs](https://mlsanity.com/paper/2404.00456/) (2024)
- [In-Datacenter Performance Analysis of a Tensor Processing Unit](https://mlsanity.com/paper/1704.04760/) (2017)
- [Documenting Large Webtext Corpora: A Case Study on the Colossal Clean Crawled Corpus](https://mlsanity.com/paper/2104.08758/) (2021)
- [EIE: Efficient Inference Engine on Compressed Deep Neural Network](https://mlsanity.com/paper/1602.01528/) (2016)
- [TPU v4: An Optically Reconfigurable Supercomputer for Machine Learning with Hardware Support for Embeddings](https://mlsanity.com/paper/2304.01433/) (2023)
- [Microscaling Data Formats for Deep Learning](https://mlsanity.com/paper/2310.10537/) (2023)
- [OliVe: Accelerating Large Language Models via Hardware-friendly Outlier-Victim Pair Quantization](https://mlsanity.com/paper/2304.07493/) (2023)
- [EdgeBERT: Sentence-Level Energy Optimizations for Latency-Aware Multi-Task NLP Inference](https://mlsanity.com/paper/2011.14203/) (2020)
- [LLM-FP4: 4-Bit Floating-Point Quantized Transformers](https://mlsanity.com/paper/2310.16836/) (2023)
- [VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision Neural Network Inference](https://mlsanity.com/paper/2102.04503/) (2021)
- [CirCNN: Accelerating and Compressing Deep Neural Networks Using Block-CirculantWeight Matrices](https://mlsanity.com/paper/1708.08917/) (2017)
- [ANT: Exploiting Adaptive Numerical Data Type for Low-bit Deep Neural Network Quantization](https://mlsanity.com/paper/2208.14286/) (2022)
- [With Shared Microexponents, A Little Shifting Goes a Long Way](https://mlsanity.com/paper/2302.08007/) (2023)
- [Training DNNs with Hybrid Block Floating Point](https://mlsanity.com/paper/1804.01526/) (2018)
- [FAST: DNN Training Under Variable Precision Block Floating Point with Stochastic Rounding](https://mlsanity.com/paper/2110.15456/) (2021)
- [Tender: Accelerating Large Language Models via Tensor Decomposition and Runtime Requantization](https://mlsanity.com/paper/2406.12930/) (2024)
- [Mokey: Enabling Narrow Fixed-Point Inference for Out-of-the-Box Floating-Point Transformer Models](https://mlsanity.com/paper/2203.12758/) (2022)
- [Dissecting the NVidia Turing T4 GPU via Microbenchmarking](https://mlsanity.com/paper/1903.07486/) (2019)
- [Modeling Deep Learning Accelerator Enabled GPUs](https://mlsanity.com/paper/1811.08309/) (2018)
- [G10: Enabling An Efficient Unified GPU Memory and Storage Architecture with Smart Tensor Migrations](https://mlsanity.com/paper/2310.09443/) (2023)

## Built on by (most cited first, 11 of 11)

- [M2XFP: A Metadata-Augmented Microscaling Data Format for Efficient Low-bit Quantization](https://mlsanity.com/paper/2601.19213/) (2026)
- [BATQuant: Outlier-resilient MXFP4 Quantization via Learnable Block-wise Optimization](https://mlsanity.com/paper/2603.16590/) (2026)
- [P3-LLM: An Integrated NPU-PIM Accelerator for Edge LLM Inference Using Hybrid Numerical Formats](https://mlsanity.com/paper/2511.06838/) (2025)
- [LATMiX: Learnable Affine Transformations for Microscaling Quantization of LLMs](https://mlsanity.com/paper/2602.17681/) (2026)
- [MF-QAT: Multi-Format Quantization-Aware Training for Elastic Inference](https://mlsanity.com/paper/2604.00529/) (2026)
- [A Motion-Aware Vector Quantization Framework with Centroid Reuse for Efficient VLA Inference](https://mlsanity.com/paper/2607.24148/) (2026)
- [GlowQ: Group-Shared LOw-Rank Approximation for Quantized LLMs](https://mlsanity.com/paper/2603.25385/) (2026)
- [Cassandra: Enabling Reasoning LLMs at Edge via Self-Speculative Decoding](https://mlsanity.com/paper/2605.26558/) (2026)
- [Heterogeneity-Aware Microscaling for Efficient Low-Bit LLM Inference](https://mlsanity.com/paper/2608.03867/) (2026)
- [HBQ: Hierarchical Scaling Block Quantization with Hardware-Efficiency-Aware Design for Accurate LLM Inference](https://mlsanity.com/paper/2609.00450/) (2026)
- [MiX: Micro-Inverted-Scaling for End-to-End Low-Bit Vision-Language Model Acceleration](https://mlsanity.com/paper/2609.19683/) (2026)

## Abstract (arXiv)

Reduced-precision data formats are crucial for cost-effective serving of large language models (LLMs). While numerous reduced-precision formats have been introduced thus far, they often require intrusive modifications to the software frameworks or are rather unconventional for widespread adoption across hardware vendors. In this paper, we instead focus on recent industry-driven variants of block floating-point (BFP) formats and conduct a comprehensive analysis to push their limits for efficient LLM serving. Our analysis shows that existing ultra low-bit BFP variants struggle to provide reasonable language model performance due to outlier values in blocks. To address the outliers with BFPs, we propose MX+, a cost-effective and non-intrusive extension designed for seamless integration into the microscaling (MX) formats. MX+ builds on the key insight that the outlier does not need to use its exponent field in the element data type, which allows us to repurpose the exponent field as an extended mantissa to increase the precision of the outlier element. Our evaluation shows that MX+ achieves significantly higher model performance compared to the 4-bit MX format (MXFP4) with negligible storage overhead and slowdown, thus offering a compelling alternative to MXFP4 or MXFP6 for efficient LLM inference.