Language Processing Unit, every mention
7 scenes across 2 shows · ← back to Language Processing Unit
tap a year for its mentions
20VC 7
David Senra 3
every year every show 20VC 7 David Senra 3
Verbatim, from the transcripts: passages where Language Processing Unit comes up on 20VC, David Senra
Why Asking the Right Questions Is the Most Important Skill in the AI Age
- ▶ 29:19 Jonathan Ross The difference is you put, you know, you put these LPUs into a system and all of a sudden the generation of tokens gets faster. 3 times in the scene
Groq Founder, Jonathan Ross: OpenAI & Anthropic Will Build Their Own Chips & Will NVIDIA Hit $10TRN · 20VC with Harry Stebbings
- ▶ 44:16 Jonathan Ross And even if we end up, um, supplying 10 times as many LPUs as GPUs, all that's gonna do is increase the demand for GPUs and allow them to charge an even higher margin.
Jonathan Ross, Founder & CEO @ Groq: NVIDIA vs Groq - The Future of Training vs Inference | E1260 · 20VC with Harry Stebbings
- ▶ 15:21 Jonathan Ross I think the simplest way to understand it is equate an LPU or a GPU to an employee, right? 2 times in the scene
- ▶ 19:28 Jonathan Ross LPUs, very different.
- ▶ 28:59 Jonathan Ross But if you want lower cost and faster, then you want an LPU.
- ▶ 37:33 Jonathan Ross A model has a pretty specific use, but accelerators like LPUs and GPUs can be used for other things besides generative AI or LLMs.
- ▶ 55:12 Jonathan Ross We scaled from 620 or 640 LPUs last year at the beginning to 40,000.