# UCNN: Exploiting Computational Reuse in Deep Neural Networks via Weight Repetition

arXiv: https://arxiv.org/abs/1804.06508 · HTML: https://mlsanity.com/paper/1804.06508/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Apr 2018.

## Built on by (most cited first, 20 of 20)

- [Mind Mappings: Enabling Efficient Algorithm-Accelerator Mapping Space Search](https://mlsanity.com/paper/2103.01489/) (2021)
- [A Survey on Deep Learning Hardware Accelerators for Heterogeneous HPC Platforms](https://mlsanity.com/paper/2306.15552/) (2023)
- [Resource-Efficient Deep Learning: A Survey on Model-, Arithmetic-, and Implementation-Level Techniques](https://mlsanity.com/paper/2112.15131/) (2021)
- [LoAS: Fully Temporal-Parallel Dataflow for Dual-Sparse Spiking Neural Networks](https://mlsanity.com/paper/2407.14073/) (2024)
- [Survey: Exploiting Data Redundancy for Optimization of Deep Learning](https://mlsanity.com/paper/2208.13363/) (2022)
- [FSL-HDnn: A 5.7 TOPS/W End-to-end Few-shot Learning Classifier Accelerator with Feature Extraction and Hyperdimensional Computing](https://mlsanity.com/paper/2409.10918/) (2024)
- [MCBP: A Memory-Compute Efficient LLM Inference Accelerator Leveraging Bit-Slice-enabled Sparsity and Repetitiveness](https://mlsanity.com/paper/2509.10372/) (2025)
- [MERCURY: Accelerating DNN Training By Exploiting Input Similarity](https://mlsanity.com/paper/2110.14904/) (2021)
- [Retrospective: EIE: Efficient Inference Engine on Sparse and Compressed Neural Network](https://mlsanity.com/paper/2306.09552/) (2023)
- [CREW: Computation Reuse and Efficient Weight Storage for Hardware-accelerated MLPs and RNNs](https://mlsanity.com/paper/2107.09408/) (2021)
- [E^2TAD: An Energy-Efficient Tracking-based Action Detector](https://mlsanity.com/paper/2204.04416/) (2022)
- [Signed Binary Weight Networks](https://mlsanity.com/paper/2211.13838/) (2022)
- [Exploiting Beam Search Confidence for Energy-Efficient Speech Recognition](https://mlsanity.com/paper/2101.09083/) (2021)
- [Optimizing the Whole-life Cost in End-to-end CNN Acceleration](https://mlsanity.com/paper/2104.05541/) (2021)
- [CoDR: Computation and Data Reuse Aware CNN Accelerator](https://mlsanity.com/paper/2104.09798/) (2021)
- [Deep Neural Networks Based Weight Approximation and Computation Reuse for 2-D Image Classification](https://mlsanity.com/paper/2105.02954/) (2021)
- [PLUM: Improving Inference Efficiency By Leveraging Repetition-Sparsity Trade-Off](https://mlsanity.com/paper/2312.01581/) (2023)
- [Hardware-software co-exploration with racetrack memory based in-memory computing for CNN inference in embedded systems](https://mlsanity.com/paper/2507.01429/) (2025)
- [Reducing the Complexity of Deep Learning Models for EEG Analysis on Wearable Devices](https://mlsanity.com/paper/2606.12742/) (2026)
- [DeVIT: Low-Power Vision Transformer Acceleration Using Delta Computation](https://mlsanity.com/paper/2608.01343/) (2026)

## Abstract (arXiv)

Convolutional Neural Networks (CNNs) have begun to permeate all corners of electronic society (from voice recognition to scene generation) due to their high accuracy and machine efficiency per operation. At their core, CNN computations are made up of multi-dimensional dot products between weight and input vectors. This paper studies how weight repetition ---when the same weight occurs multiple times in or across weight vectors--- can be exploited to save energy and improve performance during CNN inference. This generalizes a popular line of work to improve efficiency from CNN weight sparsity, as reducing computation due to repeated zero weights is a special case of reducing computation due to repeated weights. To exploit weight repetition, this paper proposes a new CNN accelerator called the Unique Weight CNN Accelerator (UCNN). UCNN uses weight repetition to reuse CNN sub-computations (e.g., dot products) and to reduce CNN model size when stored in off-chip DRAM ---both of which save energy. UCNN further improves performance by exploiting sparsity in weights. We evaluate UCNN with an accelerator-level cycle and energy model and with an RTL implementation of the UCNN processing element. On three contemporary CNNs, UCNN improves throughput-normalized energy consumption by 1.2x - 4x, relative to a similarly provisioned baseline accelerator that uses Eyeriss-style sparsity optimizations. At the same time, the UCNN processing element adds only 17-24% area overhead relative to the same baseline.