Celty: SpMspV GPU Kernel and SIMT Co-Design for Efficient Dual-Sparse LLM Inference

• URL: http://arxiv.org/abs/2608.01536v1
• Date: Sun, 02 Aug 2026 23:10:04 GMT
• Title: Celty: SpMspV GPU Kernel and SIMT Co-Design for Efficient Dual-Sparse LLM Inference
• Abstract summary: Celty is a co-designed sparse format, GPU kernel, and SIMT microarchitecture for efficient spMspV in LLM inference.<n>The Celty GPU kernel achieves up to 2.8x speedup over cuBLAS and 2.4x over Flash-LLM.<n>With the Sparse SIMT Core, speedups reach up to 5.3x over cuBLAS at 70% dual-sparsity.
• Score: 10.864469045553117
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large Language Models (LLMs) increasingly rely on sparsity to reduce inference cost, but most prior work targets a single sparsity source-either weight or activation-and optimizes for batched multi-user inference. Dual-sparsity, which combines unstructured weight pruning with runtime activation sparsity, offers a compelling tradeoff among model size, accuracy, and latency for single-user decoding, but formulates as a Sparse Matrix-Sparse Vector (spMspV) workload that existing GPU kernels handle poorly. We propose Celty, a co-designed sparse format, GPU kernel, and SIMT microarchitecture for efficient spMspV in LLM inference. At the kernel level, Celty introduces a Run-Length Compressed CSC (RLC-CSC) format that enables vectorized loading of compressed weight columns and exploits both sparsity sources to skip unnecessary memory accesses, with shared memory used for scattered partial-product accumulation. At the microarchitecture level, the Celty Sparse SIMT Core integrates a pipelined RLC decoder to eliminate software-level index reconstruction and repurposes local register files for conflict-free accumulation-operating directly on the same RLC-CSC format without data layout changes. The Celty GPU kernel achieves up to 2.8x speedup over cuBLAS and 2.4x over Flash-LLM. With the Sparse SIMT Core, speedups reach up to 5.3x over cuBLAS at 70% dual-sparsity.

Related papers

• Unified Static-Dynamic Pruning for Efficient LLM Inference [6.752566603284781]
  This paper presents SPDP, a unified adaptive-inference framework.<n>It integrates unstructured SP with input-scale inference on GPU.<n>It delivers substantial throughput and performance efficiency for large-scale computation.
  arXiv  Detail & Related papers  (2026-07-24T05:19:41Z)
• Accelerating GPU Inference of Large Language Models with Moderately Unstructured Sparse Weight Matrices [18.428868416628017]
  This paper proposes an efficient inference method for large language models (LLMs) with moderate sparsity.<n>It achieves up to 1.64x-level speedup over SpInfer and up to 1.41x end-to-end speedups over FlashLLM.
  arXiv  Detail & Related papers  (2026-06-13T13:38:27Z)
• Compiling Code LLMs into Lightweight Executables [18.677096682203697]
  Ditto is a novel method for optimizing both the model size of Code LLMs and their inference programs.<n>We evaluate Ditto on three popular Code LLMs, achieving up to 10.5$times$ lower memory usage compared with their original inference pipeline.
  arXiv  Detail & Related papers  (2026-03-31T14:40:19Z)
• ZipServ: Fast and Memory-Efficient LLM Inference with Hardware-Aware Lossless Compression [19.538318240352424]
  Lossless model compression holds tremendous promise for alleviating the memory and bandwidth bottlenecks in bit-exact Large Language Model (LLM) serving.<n>Existing approaches often result in substantial inference slowdowns due to fundamental design mismatches with GPU architectures.<n>We present ZipServ, a lossless compression framework co-designed for efficient LLM inference.
  arXiv  Detail & Related papers  (2026-03-18T07:21:21Z)
• PAPI: Exploiting Dynamic Parallelism in Large Language Model Decoding with a Processing-In-Memory-Enabled Computing System [13.678531084541666]
  We propose PAPI, a PIM-enabled heterogeneous architecture that exploits dynamic scheduling of compute-bound or memory-bound kernels to suitable hardware units.<n>PAPI achieves 1.8$times$ and 11.1$times$ speed over a state-of-the-art heterogeneous accelerator and a state-of-the-art PIM-only accelerator.
  arXiv  Detail & Related papers  (2025-02-21T13:52:31Z)
• A Universal Framework for Compressing Embeddings in CTR Prediction [68.27582084015044]
  We introduce a Model-agnostic Embedding Compression (MEC) framework that compresses embedding tables by quantizing pre-trained embeddings.<n>Our approach consists of two stages: first, we apply popularity-weighted regularization to balance code distribution between high- and low-frequency features.<n> Experiments on three datasets reveal that our method reduces memory usage by over 50x while maintaining or improving recommendation performance.
  arXiv  Detail & Related papers  (2025-02-21T10:12:34Z)
• Breaking the Memory Barrier: Near Infinite Batch Size Scaling for Contrastive Loss [59.835032408496545]
  We propose a tile-based strategy that partitions the contrastive loss calculation into arbitrary small blocks. We also introduce a multi-level tiling strategy to leverage the hierarchical structure of distributed systems. Compared to SOTA memory-efficient solutions, it achieves a two-order-of-magnitude reduction in memory while maintaining comparable speed.
  arXiv  Detail & Related papers  (2024-10-22T17:59:30Z)
• LoRC: Low-Rank Compression for LLMs KV Cache with a Progressive Compression Strategy [59.1298692559785]
  Key-Value ( KV) cache is crucial component in serving transformer-based autoregressive large language models (LLMs) Existing approaches to mitigate this issue include: (1) efficient attention variants integrated in upcycling stages; (2) KV cache compression at test time; and (3) KV cache compression at test time. We propose a low-rank approximation of KV weight matrices, allowing plug-in integration with existing transformer-based LLMs without model retraining. Our method is designed to function without model tuning in upcycling stages or task-specific profiling in test stages.
  arXiv  Detail & Related papers  (2024-10-04T03:10:53Z)
• MARLIN: Mixed-Precision Auto-Regressive Parallel Inference on Large Language Models [58.3342517278868]
  This paper describes the design of Mixed-precision AutoRegressive LINear kernels. It shows that batchsizes up to 16-32 can be supported with close to maximum ($4times$) quantization speedup. MarLIN accomplishes this via a combination of techniques, such as asynchronous memory access, complex task scheduling and pipelining.
  arXiv  Detail & Related papers  (2024-08-21T16:10:41Z)
• vTensor: Flexible Virtual Tensor Management for Efficient LLM Serving [53.972175896814505]
  Large Language Models (LLMs) are widely used across various domains, processing millions of daily requests. Large Language Models (LLMs) are widely used across various domains, processing millions of daily requests.
  arXiv  Detail & Related papers  (2024-07-22T14:37:58Z)
• Fast Matrix Multiplications for Lookup Table-Quantized LLMs [58.11584672945781]
  FLUTE is a flexible lookup table engine for LUT-quantized LLMs.<n>At batch sizes 32 and quantization group size of 128, the FLUTE kernel can be 2-4x faster than existing GEMM kernels.
  arXiv  Detail & Related papers  (2024-07-15T17:55:42Z)

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.