AC-Copilot: A Co-design Toolchain for AI Accelerator

ACCESS
Overview

AC-Copilot is an application-algorithm-hardware co-design toolchain for the development of deep learning accelerators.

  • AC-Copilot: A Co-design Toolchain for AI Accelerator
Commercialisation opportunities
Technology licensing
Problem addressed

AC-Copilot can help engineers automate the design of accelerator hardware architectures early in the design process and quickly generate custom compilers and software stacks.

Innovation
  • Neural Network Architecture Search Module: Fine-tune the neural network architecture for specific application areas.
  • Hardware-aware Neural Network Compression Module: Based on the feedback from the fast architecture simulator, the neural network is quantized and compressed and pruned.
  • Hardware Architecture Optimization Module: Includes accelerator architecture search, compilation, and simulation to explore the hardware design space, reduce global data access, and optimize inference performance, power, and area (PPA).
Key impact
  • AC-Copilot's neural network architecture search and hardware-aware network compression module increase inference speed without sacrificing accuracy. Through iterative search, the architecture optimization module finds the optimal architecture configuration through model compilation and hardware simulation, improving computing power and energy efficiency.
  • A 28nm commercial-grade Transformer Accelerator test chip (AC-Transformer) generated by AC-Copilot is 16.3 times more computationally efficient and 7.2 times more energy efficient than an existing product (8nm) in the inference of vision transformer.
Award
  • The 49th International Exhibition of Inventions Geneva 2024, Silver Medal
Application
  • Development of AI accelerator chips

Specialized AI chips will be a key factor driving the AI revolution.  To cope with this emerging demand on hardware, the AI Chip Center for Emerging Smart Systems (ACCESS) is putting Hong Kong on the global map of AI chip and hardware design.  ACCESS is a tightly coordinated, multi-disciplinary center for advancing integrated circuit (IC) design technologies to enable novel data-centric computing paradigms supporting a wide range of AI applications.  Putting together world-class experts, the Center is focus on the research theme of designing customised AI chips to realise ubiquitous AI applications used throughout society.

The research agenda in ACCESS is organised into four programmes:

  • Enabling Technology for Emerging Computing Systems addresses memory and data bandwidth problems to alleviate the bottlenecks of AI hardware by exploring integration of silicon-compatible emerging technologies with scaled silicon chips.
  • Architecture and Heterogeneous System Integration focuses on exploring different new architectural and system integration solutions for efficient neuromorphic computing on platforms ranging from cloud to smart Internet of Things.
  • AI-Assisted EDA (Electronic Design Automation) for AI Hardware hopes to develop new design methodologies and design automation tools for AI chips.
  • Hardware-Accelerated AI Applications emphasise selected emerging applications for hardware acceleration, exploring system architecture and new design tools specific to the target applications to achieve breakthroughs in AI hardware in speeds and energy efficiency.

For more inforamtion, please check: https://inno-access.hk/

Enquiry
Related Innovations