Hah, yes, same. But it's basically 1 GPU with multiple cores. I am more excited about the 192 Gb GPU memory here. But then, would it support mixed-precision training and bfloat 16? Probably no. So that's actually worse than than an A100 with 48 Gb for example.
AI HARDWARE
-
Apple Headset Future Discussed: Industry Leaders Await Technology
By
–
We are just finishing up three plus hours on Twitter Spaces talking to many industry pioneers about the future of the Apple headset. Thank you. So much gratitude to write in the new Apple journal. 🙂 We still are waiting. And will be waiting for years more for the technology
-
Team effort brings AI headset to market successfully
By
–
Hats off to those who went to all those meetings, fought for resources, built teams, and got this to market. Here's Sterling's report about working on the headset.
-
Building Computers for AI: Webinar with Tenstorrent VP
By
–
Today is the last day to sign up to join our VP of #AI Matthew Mattina for our 'Building Computers for AI' Webinar tomorrow at 2:30 pm ET. Find out more here –> http://
tenstorrent.com/webinar #hardware #software -
Apple’s Vision Team Contradicts Face Computer Strategy for Kids
By
–
Apple: our vision team is going to make sure your kids go outside and aren't too close to the iPad Also Apple: face computer go brrr
-
Apple Vision Pro: Expensive Now, Cheaper Versions Coming Soon
By
–
And here is the Apple Vision Pro: cool tech but expensive too. Expect lower cost non-Pro versions in the future.
-
High LLM Costs Hinder Quality and AI Revolution
By
–
The enormous costs of running #LLMs like #ChatGPT and #Bard are limiting their quality and the global #AI revolution. Chip scarcity is also a constraint, pushing even #BigTech to monetize #chatbots prematurely. #GenerativeAI http://ow.ly/Orz250OG79E
-
Generative AI and Vision Pro: an explosive combination
By
–
Generative AI + Apple’s Vision Pro headset will be wild. Layer Stable Diffusion variations of what’s in front of you.
-
Apple’s Transformer Model for Autocorrect on Neural Engine
By
–
Autocorrect/Dictation now leverages a transformer-based model using Apple Silicon/Neural Engine… with no info on how it's inferenced. Is this an Apple-based LLM from scratch or something fine-tuned? Or something diff altogether? (Knowing the company it's probably the former.)