Running ASR for smart homes in the NPU of Intel processors
The article discusses the author's experience running automatic speech recognition (ASR) for smart homes using the NPU of Intel processors. The author upgraded their home server and explored the performance of the Intel Core Ultra 7 265K's built-in NPU for ASR tasks. The results showed significant improvements in latency and energy efficiency compared to traditional CPU processing.
- ▪The author transitioned from using an AMD GPU to an Intel Core Ultra 7 265K for ASR tasks.
- ▪Using the NPU resulted in 3-6 times faster processing and 10-22 times less energy consumption per transcription.
- ▪The author packaged their solution into a Docker image for easier deployment.
Opening excerpt (first ~120 words) tap to expand
try { if(localStorage) { let currentUser = localStorage.getItem('current_user'); if (currentUser) { currentUser = JSON.parse(currentUser); if (currentUser.id === 3951114) { document.getElementById('article-show-container').classList.add('current-user-is-article-author'); } } } } catch (e) { console.error(e); } Miguel Camba Posted on May 25 Running ASR for smart homes in the NPU of Intel processors #machinelearning #smarthome I run my own smart home — Home Assistant, voice assistant pipeline, the whole self-hosted thing. The speech-to-text step (Parakeet TDT 0.6B v3 over the Wyoming protocol) had been running on my i3 1220P intel NUC with an 12gb RTX 3060 eGPU for months.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at DEV.to (Top).
Discussion
0 commentsMore from DEV.to (Top)
