# Primitive-Driven Acceleration of Hyperdimensional Computing for Real-Time Image Classification

arXiv: https://arxiv.org/abs/2601.20061 · HTML: https://mlsanity.com/paper/2601.20061/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 37 references that are in the map.
- First posted on arXiv 27 Jan 2026.

## Builds on (most cited first, 11 of 11)

- [Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms](https://mlsanity.com/paper/1708.07747/) (2017)
- [A Survey on Hyperdimensional Computing aka Vector Symbolic Architectures, Part I: Models and Data Transformations](https://mlsanity.com/paper/2111.06077/) (2021)
- [A Survey on Hyperdimensional Computing aka Vector Symbolic Architectures, Part II: Applications, Cognitive Models, and Challenges](https://mlsanity.com/paper/2112.15424/) (2021)
- [Hyperdimensional computing as a framework for systematic aggregation of image descriptors](https://mlsanity.com/paper/2101.07720/) (2021)
- [LeHDC: Learning-Based Hyperdimensional Computing Classifier](https://mlsanity.com/paper/2203.09680/) (2022)
- [Hypervector Design for Efficient Hyperdimensional Computing on Edge Devices](https://mlsanity.com/paper/2103.06709/) (2021)
- [HDTorch: Accelerating Hyperdimensional Computing with GP-GPUs for Design Space Exploration](https://mlsanity.com/paper/2206.04746/) (2022)
- [An Encoding Framework for Binarized Images using HyperDimensional Computing](https://mlsanity.com/paper/2312.00454/) (2023)
- [Laplace-HDC: Understanding the geometry of binary hyperdimensional computing](https://mlsanity.com/paper/2404.10759/) (2024)
- [HPVM-HDC: A Heterogeneous Programming System for Accelerating Hyperdimensional Computing](https://mlsanity.com/paper/2410.15179/) (2024)
- [Hardware-Aware Static Optimization of Hyperdimensional Computations](https://mlsanity.com/paper/2304.03335/) (2023)

## Abstract (arXiv)

Hyperdimensional Computing (HDC) represents data using extremely high-dimensional, low-precision vectors, termed hypervectors (HVs), and performs learning and inference through lightweight, noise-tolerant operations. However, the high dimensionality, sparsity, and repeated data movement involved in HDC make these computations difficult to accelerate efficiently on conventional processors. As a result, executing core HDC operations: binding, permutation, bundling, and similarity search: on CPUs or GPUs often leads to suboptimal utilization, memory bottlenecks, and limits on real-time performance. In this paper, our contributions are two-fold. First, we develop an image-encoding algorithm that, similar in spirit to convolutional neural networks, maps local image patches to hypervectors enriched with spatial information. These patch-level hypervectors are then merged into a global representation using the fundamental HDC operations, enabling spatially sensitive and robust image encoding. This encoder achieves 95.67% accuracy on MNIST and 85.14% on Fashion-MNIST, outperforming prior HDC-based image encoders. Second, we design an end-to-end accelerator that implements these compute operations on an FPGA through a pipelined architecture that exploits parallelism both across the hypervector dimensionality and across the set of image patches. Our Alveo U280 implementation delivers 0.09ms inference latency, achieving up to 1300x and 60x speedup over state-of-the-art CPU and GPU baselines, respectively.