Reducing Peak Memory Usage for Modern Multimodal Large Language Model Pipelines

• URL: http://arxiv.org/abs/2604.16734v1
• Date: Fri, 17 Apr 2026 22:45:57 GMT
• Title: Reducing Peak Memory Usage for Modern Multimodal Large Language Model Pipelines
• Abstract summary: Multimodal large language models (MLLMs) have recently demonstrated strong capabilities in understanding and generating responses from diverse visual inputs.<n>As these models scale to richer visual representations, inference increasingly relies on storing large numbers of vision tokens in the key-value cache.<n>We show that MLLMs exhibit inherent structural regularities and representational redundancy that can be exploited to control memory growth throughout inference.
• Score: 4.564896252406974
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Multimodal large language models (MLLMs) have recently demonstrated strong capabilities in understanding and generating responses from diverse visual inputs, including high-resolution images and long video sequences. As these models scale to richer visual representations, inference increasingly relies on storing large numbers of vision tokens in the key-value (KV) cache, making memory consumption a central bottleneck. Existing methods address this issue by identifying redundancy in vision tokens and compressing the cache, but such compression is typically applied only after all inputs are processed, resulting in high peak memory usage during the prefill stage. In this work, we show that MLLMs exhibit inherent structural regularities and representational redundancy that can be exploited to control memory growth throughout inference. Based on this insight, we propose a sequential input-compression mechanism that enforces a fixed memory budget by performing structure-aware key-value cache compression during the prefill process. This approach substantially reduces peak memory usage while maintaining generative performance with only minimal degradation, enabling more practical and memory-efficient multimodal inference.

Related papers

• KVCapsule: Efficient Sequential KV Cache Compression for Vision-Language Models with Asymmetric Redundancy [8.24188371483346]
  Vision-Language Models (VLMs) have emerged as a critical and fast-growing extension of Large Language Models (LLMs)<n>We propose KVCapsule, a novel KV cache compression framework for vision tokens.<n>We demonstrate up to 2x improvement in TPS and 2.4x reduction in KV cache memory at a 60% compression ratio, with negligible degradation in accuracy or response quality.
  arXiv  Detail & Related papers  (2026-05-14T23:01:58Z)
• ZoomR: Memory Efficient Reasoning through Multi-Granularity Key Value Retrieval [58.575695990976136]
  Large language models (LLMs) have shown great performance on complex reasoning tasks.<n>LLMs often require generating long intermediate thoughts before reaching a final answer.<n>We introduce ZoomR, a novel approach that enables LLMs to adaptively compress verbose reasoning thoughts into summaries.
  arXiv  Detail & Related papers  (2026-04-13T02:00:35Z)
• CalibQuant: 1-Bit KV Cache Quantization for Multimodal LLMs [45.77132019859689]
  CalibQuant is a visual quantization strategy that drastically reduces both memory and computational overhead.<n>We achieve a 10x throughput increase on InternVL models.
  arXiv  Detail & Related papers  (2025-02-15T05:08:01Z)
• CSR:Achieving 1 Bit Key-Value Cache via Sparse Representation [63.65323577445951]
  We propose a novel approach called Cache Sparse Representation (CSR)<n>CSR transforms the dense Key-Value cache tensor into sparse indexes and weights, offering a more memory-efficient representation during LLM inference.<n>Our experiments demonstrate CSR achieves performance comparable to state-of-the-art KV cache quantization algorithms.
  arXiv  Detail & Related papers  (2024-12-16T13:01:53Z)
• A Method for Building Large Language Models with Predefined KV Cache Capacity [11.710667043543545]
  The Bounded-Cache Transformer (BCT) addresses the excessive memory consumption issue in traditional KV caches.<n>By dynamically updating the key-value vector sequences, the BCT achieves efficient inference within limited cache capacity.<n> Experimental results demonstrate that the BCT significantly reduces memory usage while maintaining the model's inference quality.
  arXiv  Detail & Related papers  (2024-11-24T11:30:00Z)
• ThinK: Thinner Key Cache by Query-Driven Pruning [63.13363917871414]
  Large Language Models (LLMs) have revolutionized the field of natural language processing, achieving unprecedented performance across a variety of applications.<n>This paper focuses on the long-context scenario, addressing the inefficiencies in KV cache memory consumption during inference.<n>We propose ThinK, a novel query-dependent KV cache pruning method designed to minimize attention weight loss while selectively pruning the least significant channels.
  arXiv  Detail & Related papers  (2024-07-30T17:59:08Z)
• Efficient Inference of Vision Instruction-Following Models with Elastic Cache [76.44955111634545]
  We introduce Elastic Cache, a novel strategy for efficient deployment of instruction-following large vision-language models. We propose an importance-driven cache merging strategy to prune redundancy caches. For instruction encoding, we utilize the frequency to evaluate the importance of caches. Results on a range of LVLMs demonstrate that Elastic Cache not only boosts efficiency but also notably outperforms existing pruning methods in language generation.
  arXiv  Detail & Related papers  (2024-07-25T15:29:05Z)
• InfiniGen: Efficient Generative Inference of Large Language Models with Dynamic KV Cache Management [0.5899781520375794]
  Transformer-based large language models (LLMs) demonstrate impressive performance across various natural language processing tasks. serving inference for generating long contents poses a challenge due to the enormous memory footprint of the transient state. InfiniGen is a novel KV cache management framework tailored for long-text generation.
  arXiv  Detail & Related papers  (2024-06-28T07:41:26Z)
• LOOK-M: Look-Once Optimization in KV Cache for Efficient Multimodal Long-Context Inference [32.20654044142376]
  LOOK-M is a pioneering, fine-tuning-free approach that efficiently reduces the multimodal KV cache size. It achieves up to 1.5x faster decoding and also maintains or even enhances performance across a variety of long context multimodal tasks.
  arXiv  Detail & Related papers  (2024-06-26T07:44:24Z)
• CORM: Cache Optimization with Recent Message for Large Language Model Inference [57.109354287786154]
  We introduce an innovative method for optimizing the KV cache, which considerably minimizes its memory footprint. CORM, a KV cache eviction policy, dynamically retains essential key-value pairs for inference without the need for model fine-tuning. Our validation shows that CORM reduces the inference memory usage of KV cache by up to 70% with negligible performance degradation across six tasks in LongBench.
  arXiv  Detail & Related papers  (2024-04-24T16:11:54Z)
• Memformer: A Memory-Augmented Transformer for Sequence Modeling [55.780849185884996]
  We present Memformer, an efficient neural network for sequence modeling. Our model achieves linear time complexity and constant memory space complexity when processing long sequences.
  arXiv  Detail & Related papers  (2020-10-14T09:03:36Z)

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.