Subscribe
Sign in
Groq Chat
p/groq-chatAn LPU inference engine
Start new thread
Chris Messina

Groq® - Hyperfast LLM running on custom built GPUs

by
Top Hunter
•

An LPU Inference Engine, with LPU standing for Language Processing Unit™, is a new type of end-to-end processing unit system that provides the fastest inference at ~500 tokens/second.

Add a comment

Replies

Best
Johan Steneros
It is fast, that is for sure. Where can I get more information about the chips and hardware? Is there a GPU cloud service?
Avi Basnet
This seems extremely interesting- I’m curious what you’ve seen to be the biggest use case for this LLM?
Peter Schout
Congratulations! speed/accuracy is incredible, no wonder NVDA took a dip 😯
Cuong Vu
Groq is a promising product, and I believe your detailed insights could attract even more supporters, helping people better understand its value.
Ian Nance
Man, that IS fast...Already loving it : )
Borja Soler
this will be incredible for the future of LLMs and all the products benefiting from them. super excited with all the new things that will come
ROHIT GUPTA
amazing...
Kien Nguyen
Congrat on the launch? Do you have any plan when to support custom training?
Ivan Somov
Good luck! I am really excited about this hardware stuff for LLMs!
Abhilash Chowdhary
Going to give this a try, team Groq®. Looks interesting.
Sourabh
Don't know why its ranked so low as of now, the speed is awesome. It does what it says.
Congrats team Groq® on your launch.
Junior Perassoli
It looks very promising. How can I find information on how to use the APIs?
Aris Nakos
Wow, you guys are innovating. Congratulations! I tested it out and was blown away.
Daniel Rödler
Wow, love it. We are heavily relying on LLMs and the slowness of our agents is a constant annoyance. A 14x speed-up would be a real game changer. Can't wait to see LPUs in action and at scale. Keep going!
Mona Dey
This is helpful post.thanks