https://leo06051212.github.io/tags/llm/LLMHugoBlox Kit (https://hugoblox.com)en-usSat, 29 Aug 2026 00:00:00 +0000https://leo06051212.github.io/media/icon_hu_da05098ef60dc2e7.pnghttps://leo06051212.github.io/tags/llm/https://leo06051212.github.io/blog/2026-08-29-uoa-open-day-interactive-llm-demo/Sat, 29 Aug 2026 00:00:00 +0000https://leo06051212.github.io/blog/2026-08-29-uoa-open-day-interactive-llm-demo/<p> <figure > <div class="flex justify-center "> <div class="w-full" ><img src="https://leo06051212.github.io/uploads/blog/uoa-open-day-2026-interactive-llm-demo.jpg" alt="University of Auckland Open Day 2026" loading="lazy" data-zoomable /></div> </div></figure> </p> <p>On Saturday, 29 August 2026, our research group welcomed prospective students and their families to the University of Auckland Open Day. We prepared an interactive Large Language Model (LLM) demonstration, giving visitors an opportunity to experience a conversational AI system first-hand.</p> <p>The demonstration used a GPU–FPGA prefill–decode (PD) separation design to support responsive, natural-language interaction. It provided a practical way to introduce prospective first-year students to the hardware–software co-design ideas that underpin efficient AI systems.</p> <p>Yulin Fu and Jianglan Zhu, two of my PhD students, explained the internal technical details behind the demonstration and discussed how the GPU and FPGA components work together. It was a pleasure to share our research with future University of Auckland students and their families.</p>