LLM Inference Beyond a Single Node: From Bottlenecks to Mitigations with Fast All-Reduce Communication

• URL: http://arxiv.org/abs/2511.09557v2
• Date: Fri, 14 Nov 2025 02:01:08 GMT
• Title: LLM Inference Beyond a Single Node: From Bottlenecks to Mitigations with Fast All-Reduce Communication
• Authors: Prajwal Singhania, Siddharth Singh, Lannie Dalton Hough, Akarsh Srivastava, Harshitha Menon, Charles Fredrick Jekel, Abhinav Bhatele,
• Abstract summary: We present a detailed performance study of multi-node distributed inference using large language models (LLMs) on GPU-based supercomputers.<n>We conduct experiments with several state-of-the-art inference engines alongside YALIS, a research-oriented prototype engine designed for controlled experimentation.
• Score: 5.468224958799568
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: As large language models (LLMs) continue to grow in size, distributed inference has become increasingly important. Model-parallel strategies must now efficiently scale not only across multiple GPUs but also across multiple nodes. In this work, we present a detailed performance study of multi-node distributed inference using LLMs on GPU-based supercomputers. We conduct experiments with several state-of-the-art inference engines alongside YALIS, a research-oriented prototype engine designed for controlled experimentation. We analyze the strong-scaling behavior of different model-parallel schemes and identify key bottlenecks. Since all-reduce operations are a common performance bottleneck, we develop NVRAR, a hierarchical all-reduce algorithm based on recursive doubling with NVSHMEM. NVRAR achieves up to 1.9x-3.6x lower latency than NCCL for message sizes between 128 KB and 2 MB on HPE Slingshot and InfiniBand interconnects. Integrated into YALIS, NVRAR achieves up to a 1.72x reduction in end-to-end batch latency for the Llama 3.1 405B model in multi-node decode-heavy workloads using tensor parallelism.

Related papers

• Eliminating Multi-GPU Performance Taxes: A Systems Approach to Efficient Distributed LLMs [61.953548065938385]
  We introduce the ''Three Taxes'' (Bulk Synchronous, Inter- Kernel Data Locality, and Kernel Launch Overhead) as an analytical framework.<n>We propose moving beyond the rigid BSP model to address key inefficiencies in distributed GPU execution.<n>We observe a 10-20% speedup in end-to-end latency over BSP-based approaches.
  arXiv  Detail & Related papers  (2025-11-04T01:15:44Z)
• Scalable Engine and the Performance of Different LLM Models in a SLURM based HPC architecture [3.746889836344766]
  This work elaborates on a High performance computing architecture based on Simple Linux Utility for Resource Management (SLURM)<n> Dynamic resource scheduling and seamless integration of containerized have been leveraged to manage CPU, GPU, and memory efficiently in multi-node clusters.<n>The obtained results pave ways for significantly more efficient, responsive, and fault-tolerant LLM inference on large-scale HPC infrastructures.
  arXiv  Detail & Related papers  (2025-08-25T09:11:27Z)
• Enabling High-Sparsity Foundational Llama Models with Efficient Pretraining and Deployment [56.44025052765861]
  Large language models (LLMs) have revolutionized Natural Language Processing (NLP), but their size creates computational bottlenecks. We introduce a novel approach to create accurate, sparse foundational versions of performant LLMs. We show a total speedup on CPUs for sparse-quantized LLaMA models of up to 8.6x.
  arXiv  Detail & Related papers  (2024-05-06T16:03:32Z)
• Distributed Inference and Fine-tuning of Large Language Models Over The Internet [91.00270820533272]
  Large language models (LLMs) are useful in many NLP tasks and become more capable with size. These models require high-end hardware, making them inaccessible to most researchers. We develop fault-tolerant inference algorithms and load-balancing protocols that automatically assign devices to maximize the total system throughput.
  arXiv  Detail & Related papers  (2023-12-13T18:52:49Z)
• Boosting the effective performance of massively parallel tensor network state algorithms on hybrid CPU-GPU based architectures via non-Abelian symmetries [0.0]
  Non-Abelian symmetry related tensor algebra based on Wigner-Eckhart theorem is fully detached from the conventional tensor network layer. We have achieved an order of magnitude increase in performance with respect to results reported in arXiv:2305.05581 in terms of computational complexity. Our solution has an estimated effective performance of 250-500 TFLOPS.
  arXiv  Detail & Related papers  (2023-09-23T07:49:53Z)
• SqueezeLLM: Dense-and-Sparse Quantization [80.32162537942138]
  Main bottleneck for generative inference with LLMs is memory bandwidth, rather than compute, for single batch inference. We introduce SqueezeLLM, a post-training quantization framework that enables lossless compression to ultra-low precisions of up to 3-bit. Our framework incorporates two novel ideas: (i) sensitivity-based non-uniform quantization, which searches for the optimal bit precision assignment based on second-order information; and (ii) the Dense-and-Sparse decomposition that stores outliers and sensitive weight values in an efficient sparse format.
  arXiv  Detail & Related papers  (2023-06-13T08:57:54Z)
• Tensor Slicing and Optimization for Multicore NPUs [2.670309629218727]
  This paper proposes a compiler optimization pass for Multicore NPUs, called Slicing Optimization (TSO) TSO identifies the best tensor slicing that minimizes execution time for a set of CNN models. Results show that TSO is capable of identifying the best tensor slicing that minimizes execution time for a set of CNN models.
  arXiv  Detail & Related papers  (2023-04-06T12:03:03Z)
• SWARM Parallelism: Training Large Models Can Be Surprisingly Communication-Efficient [69.61083127540776]
  Deep learning applications benefit from using large models with billions of parameters. Training these models is notoriously expensive due to the need for specialized HPC clusters. We consider alternative setups for training large models: using cheap "preemptible" instances or pooling existing resources from multiple regions.
  arXiv  Detail & Related papers  (2023-01-27T18:55:19Z)
• Lightweight and Progressively-Scalable Networks for Semantic Segmentation [100.63114424262234]
  Multi-scale learning frameworks have been regarded as a capable class of models to boost semantic segmentation. In this paper, we thoroughly analyze the design of convolutional blocks and the ways of interactions across multiple scales. We devise Lightweight and Progressively-Scalable Networks (LPS-Net) that novelly expands the network complexity in a greedy manner.
  arXiv  Detail & Related papers  (2022-07-27T16:00:28Z)

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.