#Groq was thrilled to be the first stop on the @ITPresstTour so we could speak about "the most consequential AI company you haven’t heard of” (per @chamath
) & wow everyone with our demo (
http://
chat.groq.com). Thanks for coming @CDP_FST @ctsmithiii @editingwhiz @VMblog
@groqinc
-
Groq First Stop on IT Press Tour Showcases AI Demo
By
–
-
Groq LPU Offers 300T/s Speed Without Latency Issues
By
–
You know you could have access to 300T/s today, @SamA
. No need to raise money and build your own factory. The #Groq ® #LPU ™ is fabbed in the US and doesn't have the annoying latency limitations that are currently holding #ChatGPT back. @SiliconANGLE #genAI @openAI -
Groq Accelerates Generative AI for Immersive Interactive Experiences
By
–
At Groq, we aim to create a sense of awe by accelerating generative AI applications to the point that they become immersive experiences. We're thrilled to have Mark Vallis on our team as we enable truly interactive engagement with AI. pic.twitter.com/g1ckVh85BJ
— Groq Inc (@GroqInc) 22 janvier 2024At Groq, we aim to create a sense of awe by accelerating generative AI applications to the point that they become immersive experiences. We're thrilled to have Mark Vallis on our team as we enable truly interactive engagement with AI.
-
GenAI Becomes Real: Quantum Computing for AI Training
By
–
2024 is the year that #GenAI becomes real & interactive. Don't miss out on this opportunity to learn about the future of AI. Join Oskar Mencer at the Quantum Computing in Finance Conference today to hear how QUBOs & quantum computers could be used to train AI.
-
Groq Shares Monarch Matrices and Structured Matrices Presentation
By
–
Talent density is core to #Groq. We try to think differently & reimagine efficiency. We're excited to share @cdsmithus
's presentation on Monarch Matrices & Structured Matrices, part of our Lunch & Learn playlist featuring Groqsters sharing their wisdom. https://
youtu.be/Pr6nNuGSbCE -
Groq Demonstrates Ultra-Fast AI Image Processing with StyleClip GAN
By
–
"can @GroqInc also do AI Image Processing?" https://
youtu.be/E7TB5fgA8wk
Yes we do, fast! Language is more than words, like audio, images etc. See how fast our low latency is on a #GAN model #StyleClip. Don't blink, you'll miss it. #groqspeed #GenAI #RealTime http://
chat.groq.com -
Groq and AIXplain Event: Real-Time AI Inference for Consumer Electronics
By
–
Join #Groq & @aixplain for cocktails and a peek at how real-time inference can be a game changer in consumer electronics. We'll chat about the future of AI, strategies you’re considering and the new standard for AI application speed going beyond real-time.
Reach out to -
Groq’s Chat Platform Achieves 300 Tokens Per Second with Llama2
By
–
…and that open-source list means you could learn about the samely named model but on different platforms, yet it will perform differently. This is challenging for providers also. http://
Chat.Groq.com is the fastest for Llama2, 70B at ~300 tokens per second per user but that -
Groq Hosts AI Compute Cocktail Party at CES 2024
By
–
Some Grogsters, & CEO @JonathanRoss321
, will be @CES next week hosting a cocktail party. Will you be there? Want to come hang out and chat about the future of AI compute and how to ready your org? DM us and maybe we'll send you the invite form. 😉 #betterongroq #LLMs #AI #CES24 -
Groq LPU Engine: Kernel-free Llama 2 70B Inference Performance
By
–
Kernel-free, no Cuda, Compiler-only solution.
The current implementation, running on our LPU Inference Engine uses our own processor. The model is Llama 2, 70B 4k sequence at FP16…We haven't even hit the next gear through those other methods. Plenty in the tank left!