Perplexity’s Post

Accelerate your AI initiatives with Pure+NVIDIA

🔍 Ever wondered about the power of Llama 3 inference on an ARM single board computer? Wonder no more! 🦙💻 🌟 Introducing my latest exploration into the world of AI and edge computing: Llama 3 is running on these compact devices! Yes, you read that right – an 8B LLM doing its magic right in the palm of your hand! 🎩✨ 🔧 With just a sprinkle of tinkering using the incredible existing tooling, the potential for on-device ChatGPT-like interactions is now unlocked. 🗣️💬 💡 While progress may be humble at this stage, the results are already sparking curiosity and paving the way for endless possibilities. 🌈 From enhancing efficiency to igniting new project ideas, the journey has only just begun! 📣 Your feedback, insights, and ideas are invaluable as the boundaries of AI at the edge continues to be pushed. 💡💼 #AI #EdgeComputing #Llama3 #ARM #Innovation #FeedbackWelcome

Nvidia just publised a curated collection of utilities to build QA-RAG like solutions, including datasets and models weights. It is particularly important to point out the multi-turn encoder, that highlights the importance of short term memory during information retrieval step. Instead of using your LLM (or a smaller version) to reformulate your query, simply ingest your whole short-term memory in the encoder. Not only does it drastically reduce latency, but it also seems to be a much more natural approach. Kudos to NVIDIA AI team! https://lnkd.in/dRj3eVMU

Try out the new 405B parameter llama 3.1 with NVIDIA NIMs now in preview mode. https://lnkd.in/eXBWuNTd

🚀 Exciting News Alert! 🚀 Unlocking the power of LLM models just got easier with NVIDIA's groundbreaking program! 🌟 🔍 Dive into my latest YouTube video where I reveal how you can access ALL LLM models for FREE using the NVIDIA Build Program. Plus, get a whopping 1000 API credits on the house! If you're serious about leveraging cutting-edge technology to revolutionize your projects, this is a game-changer you can't afford to miss. Click the link below to watch now and unleash the full potential of AI: https://lnkd.in/gN-HvYJJ #NVIDIA #LLMModels #AI #Innovation #TechTrends #YouTubeTutorial #FreeAPICredits

Wow, it is truly astounding how computing science has evolved. I always believed that solving real-world problems, such as controlling the movements of airplanes or rockets, required exact arithmetic units with zero error tolerance. However, the advent of AI has completely upended my understanding. With FP4, how much precision can one achieve? Shockingly, still a leading AI chip design company has introduced an FP4-based chip "Blackwell" , that achieves 20,000 TFLOPs. thanks Linas Beliūnas for sharing this summary :-) #AICHIPs #FP4 #20,000TFLOPs

The future of computing is bright!!! If we use #photons in #classical_computing and #quantum_computing: optical computing paradigm. Check out #Lightmatter: Photonic AI accelerator.

Elon Musk's Groq Surprises with Strong Cost - Speed Ratio Groq was founded by Musk to compete against OPENAI and others on the strongly growing LLM market. Now, a study was conducted to compare language models based on infecence cost and speed, focusing on text generation about Nvidia's business from its 10-K report. The methodology included 10 trials per model, limiting outputs to 1000 tokens, aiming to find cost-effective and speedy solutions. Groq surprisingly emerged as an excellent performer. And while one may argue that different test conditions may result in a different results, it is a hint that Groq may be a model to be taken serious moving forward. This efficiency makes Groq a potential candidate e. g. into RAG pipelines, highlighting its ability to enhance text generation tasks while maintaining low operational costs. #AI #TechEfficiency #Groq

Upcoming livestream where we bash together some Phoenix and an LLM using Nx and get it running on a Fly GPU. I figure plenty of people want to experiment with this stuff. This should get you the fundamentals and whatever else we get up to. Thursday 21st, 1500 CET. Will be recorded. https://lnkd.in/dyhRKzdN #elixir #nx #ai #llm #ai #hashtag

Discover how AMD is helping accelerate AI in this short video!