mlsanityData to 14 Oct 2026

VEGETA: Vertically-Integrated Extensions for Sparse/Dense GEMM Tile Acceleration on CPUs

arXiv 2302.08687 ↗First posted 17 Feb 2023cs.AR cs.AI cs.LGcited by 9

Not on the map. Why?

Abstract (arXiv)

Deep Learning (DL) acceleration support in CPUs has recently gained a lot of traction, with several companies (Arm, Intel, IBM) announcing products with specialized matrix engines accessible via GEMM instructions. CPUs are pervasive and need to handle diverse requirements across DL workloads running in edge/HPC/cloud platforms. Therefore, as DL workloads embrace sparsity to reduce the computations and memory size of models, it is also imperative for CPUs to add support for sparsity to avoid under-utilization of the dense matrix engine and inefficient usage of the caches and registers. This work presents VEGETA, a set of ISA and microarchitecture extensions over dense matrix engines to support flexible structured sparsity for CPUs, enabling programmable support for diverse DL models with varying degrees of sparsity. Compared to the state-of-the-art (SOTA) dense matrix engine in CPUs, a VEGETA engine provides 1.09x, 2.20x, 3.74x, and 3.28x speed-ups when running 4:4 (dense), 2:4, 1:4, and unstructured (95%) sparse DNN layers.

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 18 of its 54 references that are in the map. First posted on arXiv 17 Feb 2023.

Builds on

18 of 54 references.

References · 18

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  3. 2020Language Models are Few-Shot Learnerscited by 32,419
  4. 2015Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Codingcited by 2,329
  5. 2015Learning both Weights and Connections for Efficient Neural Networkscited by 2,031
  6. 2017In-Datacenter Performance Analysis of a Tensor Processing Unitcited by 618
  7. 2019Deep Learning Recommendation Model for Personalization and Recommendation Systemscited by 327
  8. 2014cuDNN: Efficient Primitives for Deep Learningcited by 245
  9. 2021Accelerating Sparse Deep Neural Networkscited by 223
  10. 2021Learning N:M Fine-grained Structured Sparse Neural Networks From Scratchcited by 179
  11. 2017SCNN: An Accelerator for Compressed-sparse Convolutional Neural Networkscited by 135
  12. 2020SpArch: Efficient Architecture for Sparse Matrix Multiplicationcited by 49
6 more
  1. 2021S2TA: Exploiting Structured Sparsity for Energy-Efficient Mobile CNN Accelerationcited by 19
  2. 2020Systolic Tensor Array: An Efficient Structured-Sparse GEMM Accelerator for Mobile CNN Inferencecited by 17
  3. 2018Packing Sparse Convolutional Neural Networks for Efficient Systolic Array Implementations: Column Combining Under Joint Optimizationcited by 15
  4. 2020Fast Stencil-Code Computation on a Wafer-Scale Processorcited by 10
  5. 2017SparCE: Sparsity aware General Purpose Core Extensions to Accelerate Deep Neural Networkscited by 5
  6. 2021RASA: Efficient Register-Aware Systolic Array Matrix Engine for CPUcited by 3

Built on by

Cited by year

Citing papers per year: 2023: 1; 2024: 4; 2025: 4.

9 papers.

  1. 2024MicroScopiQ: Accelerating Foundational Models through Outlier-Aware Microscaling Quantizationcited by 28
  2. 2023Scaling Down to Scale Up: A Cost-Benefit Analysis of Replacing OpenAI's LLM with Open Source SLMs in Productioncited by 21
  3. 2024Progressive Gradient Flow for Robust N:M Sparsity Training in Transformerscited by 18
  4. 2024Enabling Unstructured Sparse Acceleration on Structured Sparse Acceleratorscited by 8
  5. 2025Accelerating LLM Inference with Flexible N:M Sparsity via A Fully Digital Compute-in-Memory Acceleratorcited by 3
  6. 2024ELSA: Exploiting Layer-wise N:M Sparsity for Vision Transformer Accelerationcited by 2
  7. 2025Lightweight Software Kernels and Hardware Extensions for Efficient Sparse Deep Neural Networks on Microcontrollerscited by 2
  8. 2025DECA: A Near-Core LLM Decompression Accelerator Grounded on a 3D Roofline Modelcited by 2
  9. 2025Periodic Online Testing for Sparse Systolic Tensor Arrayscited by 1