Hi all!
Me and my team is building an AI inference backend, especially focusing on CPU optimization for now.
It's showing great performance gain for AI models on Rpi board, even outperforming XNNPACK backend by over 2x(demo vid: https://youtu.be/u7wzFngylis)
Thanks to Rpi community, I'll be holding a webinar on how we did this:
https://events.raspberrypi.com/communit ... ef855b0b96
We've optimized some famous AI models on Rpi, which show the faster inference speed on Rpi, and will be releasing some of it during the event. So please come join us via above link, it will be held next week on Dec. 5th(Thu) 9 am PT.
Plus, let me know here if you have any AI models that you're struggling with in terms of latency - we'll consider that models to be optimized and shared with you.
Me and my team is building an AI inference backend, especially focusing on CPU optimization for now.
It's showing great performance gain for AI models on Rpi board, even outperforming XNNPACK backend by over 2x(demo vid: https://youtu.be/u7wzFngylis)
Thanks to Rpi community, I'll be holding a webinar on how we did this:
https://events.raspberrypi.com/communit ... ef855b0b96
We've optimized some famous AI models on Rpi, which show the faster inference speed on Rpi, and will be releasing some of it during the event. So please come join us via above link, it will be held next week on Dec. 5th(Thu) 9 am PT.
Plus, let me know here if you have any AI models that you're struggling with in terms of latency - we'll consider that models to be optimized and shared with you.
Statistics: Posted by ezoptim — Sat Nov 30, 2024 7:25 am — Replies 0 — Views 27