UniFormer: Unified and Efficient Transformer for Reasoning Across General and Custom Computing

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ran, Zhuoheng, Wu, Chong, Xu, Renjie, Che, Maolin, Yan, Hong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911259391164416
author Ran, Zhuoheng
Wu, Chong
Xu, Renjie
Che, Maolin
Yan, Hong
author_facet Ran, Zhuoheng
Wu, Chong
Xu, Renjie
Che, Maolin
Yan, Hong
contents The success of neural networks such as convolutional neural networks (CNNs) has been largely attributed to their effective and widespread deployment on customised computing platforms, including field-programmable gate arrays (FPGAs) and application-specific integrated circuits (ASICs). In the current era, Transformer-based architectures underpin the majority of state-of-the-art (SOTA) larger models that are also increasingly deployed on customised computing hardware for low-power and real-time applications. However, the fundamentally different parallel computation paradigms between general-purpose and customised computing often lead to compromises in model transfer and deployability, which typically come at the cost of complexity, efficiency or accuracy. Moreover, many cross-platform optimisation principles have also remained underexplored in existing studies. This paper introduces UniFormer, a unified and efficient Transformer architecture for both general-purpose and customised computing platforms. By enabling higher parallelism and compute-storage fusion, UniFormer achieves state-of-the-art (SOTA) accuracy and latency on GPUs while exhibiting strong adaptability on FPGAs. To the best of our knowledge, this paper is the first efficient Transformer work that jointly considers both general-purpose and customised computing architectures.
format Preprint
id arxiv_https___arxiv_org_abs_2511_08135
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle UniFormer: Unified and Efficient Transformer for Reasoning Across General and Custom Computing
Ran, Zhuoheng
Wu, Chong
Xu, Renjie
Che, Maolin
Yan, Hong
Distributed, Parallel, and Cluster Computing
Hardware Architecture
The success of neural networks such as convolutional neural networks (CNNs) has been largely attributed to their effective and widespread deployment on customised computing platforms, including field-programmable gate arrays (FPGAs) and application-specific integrated circuits (ASICs). In the current era, Transformer-based architectures underpin the majority of state-of-the-art (SOTA) larger models that are also increasingly deployed on customised computing hardware for low-power and real-time applications. However, the fundamentally different parallel computation paradigms between general-purpose and customised computing often lead to compromises in model transfer and deployability, which typically come at the cost of complexity, efficiency or accuracy. Moreover, many cross-platform optimisation principles have also remained underexplored in existing studies. This paper introduces UniFormer, a unified and efficient Transformer architecture for both general-purpose and customised computing platforms. By enabling higher parallelism and compute-storage fusion, UniFormer achieves state-of-the-art (SOTA) accuracy and latency on GPUs while exhibiting strong adaptability on FPGAs. To the best of our knowledge, this paper is the first efficient Transformer work that jointly considers both general-purpose and customised computing architectures.
title UniFormer: Unified and Efficient Transformer for Reasoning Across General and Custom Computing
topic Distributed, Parallel, and Cluster Computing
Hardware Architecture
url https://arxiv.org/abs/2511.08135