AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
3min chapter
Generative AI at the Edge with Vinesh Sukumar - #623
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
The Importance of Hardware Architecture for Attention-Specific Models
The focus is like speeding up matrix multiplications, right? The faster we get those, that's kind of all we need. But now that attention is all we need, are there specific structures that you see evolving in hardware to accommodate attention? Are there registers that you might build into hardware that are attention specific or kind of designed to accommodate unique properties of the transformer? Yeah, great question. When you look at these Prometheus style models, which are coming up from OpenAI team, they're like, what, 175 billion parameter kind of issues and GPT 4 is releasing this week or next week. I'm pretty sure it's north of 200 billion parameters. So the question really becomes,
00:00
Transcript
Episode notes
Get the Snipd
podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any
moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share
& Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode