David H. Lau

David H. Lau

Deep‑Tech AI — fast, local CUDA

About

I’m David — I build high‑performance, local‑first AI tools that run directly on consumer GPUs. My work focuses on CUDA acceleration, real‑time inference, and making advanced AI feel fast, private, and accessible without relying on the cloud. My goal is to create a suite of GPU‑optimized tools that give creators and developers the speed and control they’ve been missing. Whisper ASR is part of that roadmap: a push toward AI that runs locally, respects your data, and delivers performance that cloud APIs can’t match. I’m ambitious about the future of local AI. I want to help shift the industry toward tools that are faster, more private, and more user‑owned — and I’m excited to build alongside a community that values performance, transparency, and great engineering.

Badges

Tastemaker
Tastemaker
Gone streaking
Gone streaking

Forums

2h ago

Whisper ASR - Transcribe 22 min audio in 67 sec — CUDA (~20 × real‑time)

Transcribe 22 min audio in just 67 sec using the fastest GGML Whisper model (~20× real‑time). Whisper ASR is a fully local, CUDA‑accelerated transcription engine built for creators, developers, and professionals who need real‑time performance without the cloud.
View more