Multi-Scale Dequant: Eliminating Dequantization Bottleneck via Activation Decomposition for Efficient LLM Inference

• URL: http://arxiv.org/abs/2605.13915v1
• Date: Wed, 13 May 2026 09:49:56 GMT
• Title: Multi-Scale Dequant: Eliminating Dequantization Bottleneck via Activation Decomposition for Efficient LLM Inference
• Abstract summary: Quantization is essential for efficient large language model (LLM) inference.<n>Dequantization step-converting low-bit weights back to high-precision for matrix multiplication has become a critical bottleneck on modern AI accelerators.<n>This paper presents Multi-Scale Dequant (MSD), a quantization framework that removes weight/KV dequantization from the GEMM critical path.
• Score: 12.249074183271743
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Quantization is essential for efficient large language model (LLM) inference, yet the dequantization step-converting low-bit weights back to high-precision for matrix multiplication has become a critical bottleneck on modern AI accelerators. On architectures with decoupled compute units (e.g., Ascend NPUs), dequantization operations can consume more cycles than the matrix multiplication itself, leaving the high-throughput tensor cores underutilized. This paper presents Multi-Scale Dequant (MSD), a quantization framework that removes weight/KV dequantization from the GEMM critical path. Instead of lifting low-bit weights to BF16 precision, MSD decomposes high-precision BF16 activations into multiple low-precision components, each of which can be multiplied directly with quantized weights via native hardware-accelerated GEMM. This approach shifts the computational paradigm from precision conversion to multi-scale approximation, avoiding INT8-to-BF16 weight conversion before GEMM. We instantiate MSD for two weight formats and derive tight error bounds for each. For INT8 weights (W4A16), two-pass INT8 decomposition achieves near 16 effective bits. For MXFP4 weights (W4A16), two-pass MXFP4 decomposition yields near 6.6 effective bits with error bound 1/64 per block surpassing single-pass MXFP8(5.24 bits) while maintaining the same effective GEMM compute time. We further derive closed-form latency and HBM traffic models showing that MSD avoids the Vector-Cube pipeline stall caused by dequantization and reduces KV cache HBM traffic by up to 2.5 times in attention. Numerical simulations on matrix multiplication and Flash Attention kernels confirm that MSD does not degrade accuracy compared to dequantization baselines, and in many settings achieves lower L2 error.

Related papers

• TORQ: Two-Level Orthogonal Rotation for MXFP4 Quantization [13.984796236174331]
  We propose TORQ (Two-level Orthogonal Rotation for MXFP4 Quantization) as a training-free Post-Training Quantization framework.<n>We show that TORQ significantly improves the accuracy of MXFP4 activation quantization compared to existing methods.
  arXiv  Detail & Related papers  (2026-05-19T09:05:47Z)
• BWTA: Accurate and Efficient Binarized Transformer by Algorithm-Hardware Co-design [71.97035034203275]
  We analyze zero-point distortion in binarization and propose a Binary Weights & Ternary Activations (BWTA) quantization scheme.<n>For training, we propose Smooth Multi-Stage Quantization, combining a Levelwise Degradation Strategy and a MagnitudeAlignment Projection Factor.<n>Experiments show that BWTA approaches full-precision performance for Transformer-based models, with an average 3.5% drop on GLUE and less than 2% drop on five tasks.
  arXiv  Detail & Related papers  (2026-04-05T04:25:07Z)
• BPDQ: Bit-Plane Decomposition Quantization on a Variable Grid for Large Language Models [56.504879072674015]
  We propose Bit-Plane Decomposition Quantization (BPDQ), which constructs a variable quantization grid via bit-planes and scalar coefficients.<n>BPDQ enables serving Qwen2.5-72B on a single GTX 3090 with 83.85% GSM8K accuracy (vs. 90.83% at 16-bit)
  arXiv  Detail & Related papers  (2026-02-04T02:54:37Z)
• ARCQuant: Boosting NVFP4 Quantization with Augmented Residual Channels for LLMs [4.431548809730958]
  ARCQuant is a framework that boosts NVFP4 performance via Augmented Residual Channels.<n>We show that ARCQuant achieves state-of-the-art accuracy, comparable to full-precision baselines in perplexity and downstream tasks.
  arXiv  Detail & Related papers  (2026-01-12T12:27:22Z)
• FP8-Flow-MoE: A Casting-Free FP8 Recipe without Double Quantization Error [3.281844093101284]
  Training large Mixture-of-Experts (MoE) models remains computationally prohibitive due to their extreme compute and memory demands.<n>We propose FP8-Flow-MoE, a training recipe featuring a quantization-consistent FP8-centric dataflow with a scaling-aware computation and fused FP8 operators.
  arXiv  Detail & Related papers  (2025-11-04T06:36:59Z)
• AMS-QUANT: Adaptive Mantissa Sharing for Floating-point Quantization [7.413057271242686]
  Quantization, particularly floating-point quantization, is known to be capable of speeding up large language models (LLMs) inference.<n>We propose AMS-Quant, which explores floating-point quantization exploration from integer bitwidths to non-integer bit-widths.<n>We show that AMS-Quant can quantize the model to FP-5.33-e2m3 and FP4.25-e2m2, and significantly speed up the decoding over FP16 inference.
  arXiv  Detail & Related papers  (2025-10-16T15:37:23Z)
• QUAD: Quantization and Parameter-Efficient Tuning of LLM with Activation Decomposition [21.13478769431063]
  QUAD (Quantization with Activation Decomposition) is a framework leveraging Singular Value Decomposition (SVD) to suppress activation outliers for effective 4-bit quantization.<n>We show QUAD achieves 94% 96% accuracy under W4A4 quantization and 98% accuracy with W4A4/A8 and parameter-efficient fine-tuning for Llama-3 and Qwen-2.5 models.
  arXiv  Detail & Related papers  (2025-03-25T05:03:56Z)
• MergeQuant: Accurate 4-bit Static Quantization of Large Language Models by Channel-wise Calibration [23.752021919501207]
  We propose MergeQuant, an accurate and efficient per-channel static quantization framework.<n> MergeQuant integrates the per-channel quantization steps with the corresponding scalings and linear mappings through a Quantization Step Migration (QSM) method.<n>On Llama-2-7B model, MergeQuant achieves up to 1.77x speedup in decoding, and up to 2.06x speedup in end-to-end compared to FP16 baseline.
  arXiv  Detail & Related papers  (2025-03-07T04:52:28Z)
• Optimizing Large Language Model Training Using FP4 Quantization [73.55459961002371]
  Quantized training presents a promising solution by enabling low-bit arithmetic operations to reduce costs.<n>This work introduces the first FP4 training framework for large language models (LLMs)
  arXiv  Detail & Related papers  (2025-01-28T18:04:50Z)
• DB-LLM: Accurate Dual-Binarization for Efficient LLMs [83.70686728471547]
  Large language models (LLMs) have significantly advanced the field of natural language processing. Existing ultra-low-bit quantization always causes severe accuracy drops. We propose a novel Dual-Binarization method for LLMs, namely DB-LLM.
  arXiv  Detail & Related papers  (2024-02-19T09:04:30Z)
• OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models [57.27101446992148]
  Large language models (LLMs) have revolutionized natural language processing tasks. Recent post-training quantization (PTQ) methods are effective in reducing memory footprint and improving the computational efficiency of LLM. We introduce an Omnidirectionally calibrated Quantization technique for LLMs, which achieves good performance in diverse quantization settings.
  arXiv  Detail & Related papers  (2023-08-25T02:28:35Z)
• Quantized Neural Networks for Low-Precision Accumulation with Guaranteed Overflow Avoidance [68.8204255655161]
  We introduce a quantization-aware training algorithm that guarantees avoiding numerical overflow when reducing the precision of accumulators during inference. We evaluate our algorithm across multiple quantized models that we train for different tasks, showing that our approach can reduce the precision of accumulators while maintaining model accuracy with respect to a floating-point baseline.
  arXiv  Detail & Related papers  (2023-01-31T02:46:57Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.