The development of AI computing has reached a critical inflexion point. Large-language models (LLMs) have attracted tremendous attention recently and require huge computations for AI model training and inference. In the AI era, IC design now faces tremendous challenges. It’s more difficult to continuously shrink the semiconductor process to meet the targets on performance, power, area, and cost. Huge computing requirements led to high performance, bandwidth, and energy efficiency. Complex IC designs also increase the design costs and demands on highly skilled talents.
At the same time, AI also enables new opportunities. AI can create new functions in IC products and become a crucial tool to increase the design efficiency of the IC industry. Meanwhile, open-source processor architectures can provide powerful and flexible building blocks to realize complex AI computing architecture. These have brought revolutionary new opportunities for the IC industry.
On September 11th, 2024 RISC-V Taipei Day, hosted by RISC-V Taiwan Alliance (RVTA). The forum invited speakers representing local and international industry experts to share their insights on the trends and developments of RISC-V over the years. It focused on RISC-V’s role in AI Computing and explored the potential of RISC-V’s open-source architecture in the AI era; ranging from its capabilities in enhancing complex IC systems and AI computing architectures to its potential in revolutionizing computing.