Recogni

Starting 2025 right! 🚀 Our Co-Founder, Gilles Backhus, will be in London for the AI Infrastructure and Architecture Summit over the next two days. It’s a great chance to learn how our GenAI inference system can run 16-bit models at the cost of 4-bit 🤯 If you're around, feel free to say hi and chat with Gilles about our work and where GenAI is headed in 2025. See you there! 👋

2024 marked a pivotal moment in the world of AI 🚀 🔵 The explosion of larger models like Llama 405B, the debut of transformative applications like Mochi 1, and the rise of effective reasoning techniques like Chain of Thought have pushed AI inference to unprecedented heights. 🔵 Compute demand has skyrocketed, with the industry waking up to the urgent need for more efficient compute to solve the AI power provisioning problem. 🔵 A very exciting shift: realtime-oriented applications are taking center stage. The future lies in massively parallelizing models to deliver ultra-low latency, seamless user experiences. For us as a company, 2024 wasn’t just about witnessing these changes. It was about positioning ourselves to help lead them. Swipe through the slideshow to see some of the key moments from our year. Here’s to 2025.

Our Pareto Video Game was a crowd favorite at NeurIPS! We heard rave reviews like, “It’s incredibly addictive.” Now, the results are in for the top three players: 🥇 Yanxi – 18.86 bits 🥈 Radovan – 17.06 bits 🥉 Kinho – 15.54 bits Congratulations to our champions! Enjoy your vouchers 🎁

Two days in the books already 📚 Last day at NeurIPS 2024 as exhibitor. Now’s your chance to talk to us and learn more about our Pareto log math, our chip, and our system. We have great coffee, games, and our crew is ready to host you. See you there 👋

🪫 Mid-NeurIPS Recharge Alert! 🔋 Take a break from the grind and join us at NeurIPS 2024 on Thursday, December 12th, 7pm - 11pm, for an evening you won’t forget! 🤝 Great company 🍹 Good vibes: music, drinks, and food 🎓 Friends from research, academia, and industry Don’t miss this perfect opportunity to unwind and connect. We’re looking forward to seeing you there! 👉 Save your spot: https://lu.ma/208hm6rz

A new perspective on GenAI efficiency and trust Join us at EE Times AI Everywhere Online Conference, where Gilles Backhus, Co-Founder and VP of AI at Recogni, will discuss a critical challenge in the enterprise adoption of generative AI and how Recogni’s Pareto approach can help bridge the gap between trust and affordability. 📅 Date: December 11th ⏰ Time: 10:15 AM PST / 19:15 CET 💡 Session Title: Recogni Pareto: How an Increased Trust in GenAI Quality Could Unlock Enterprise Adoption As generative AI models grow in size and capability, enterprises face a dilemma: 🔵 Stick with high-precision models and bear the costs, or 🔵 Quantize for efficiency and risk a loss in quality that’s hard to measure objectively. 🔵 According to Semianalysis: “quality loss in quantized models is drastic” Gilles will explore how Recogni’s Pareto logarithmic math addresses this challenge, enabling: ✅ True 16-bit precision ✅ 4-bit operational costs ✅ Trust in results without compromise Whether you’re optimizing AI performance for software development, financial trading, or enterprise-scale applications, this session will provide a fresh take on balancing cost and precision with new mathematical approaches. 🔗 Register here: https://lnkd.in/dF7Wuxya

Recogni is heading to NeurIPS 2024! Join us in beautiful Vancouver, 🇨🇦, from December 10th to 15th, and let's connect! 📍 Find us at Booth #419, where our AI engineers will be on hand to share insights about our groundbreaking Pareto Log Math Number System and the GenAI inference revolution we’re driving at Recogni. Why stop by? Here’s what awaits you: 🎁 Recogni Pareto merchandise 🎮 A chance to win in our interactive Pareto video game ☕ Complimentary hot beverages to keep you warm We can't wait to see you there!

Today, we are excited to finally be able to share our collaboration with and investment from Juniper Networks. At Recogni, we have made it our mission to make GenAI inference profitable, sustainable, and accurate. That’s why we are developing chips and systems based on our proprietary logarithmic math number system, Pareto. And we collaborate in this mission with Juniper, whose expertise as a leader in secure, AI-Native Networking makes them a natural fit. The result? A multimodal GenAI inference system that runs the world's largest GenAI models at uncompromised 16-bit accuracy while operating at a power efficiency equivalent to 4-bit. In other words: Much cheaper GenAI production at scale, unlocking radically new applications, such as real-time AI video generation and AI agents for education and health. And this is just the beginning! Link to the press release in the comments 🔽

“Training models is a cost center, but inference is a profit center, and unless you make money on inference, ubiquitous AI is not going to happen.” Here's how we deliver data center scale GenAI inference: https://lnkd.in/dnNAG3bg

AI Hardware & Edge AI Summit, here we come 🫶 Our CPO, R K Anand will host a panel discussion with top panelists: Gaia Bellone, Alex Pham, and Michael Stewart. If you are at the conference and you want to learn more about Recogni, let us know. We are happy to connect and chat. See you there 👋