SambaNova Systems

What an incredible experience at Evident! Rodrigo Liang joined other senior leaders in the #AI & banking sector to dive into the evolving role of AI. He went on to talk more about the elements of #AgenticAI - “We built this platform that allows our customers to use open-source models and bring in the best of the best out there and then use them for their private applications. … It allows you to bring in hundreds or thousands of different agents. … And you can put them in one platform. You can’t do that on any other platform today." Thank you to the other panelist for an amazing conversation! ??? Patrick Stokes?of Salesforce ??? Tony Kim of BlackRock ??? Jacqueline O'Flanagan of Microsoft ??? Jackie Davalos of Bloomberg News

"Developers have an incredible opportunity, right here, right now, to start building off of some amazing technology, amazing models, amazing hardware." ?? Vasanth Mohan, shares why this is such an exciting time for the developer community. Start developing on the fastest AI platform on the best models out there ?? cloud.sambanova.ai #Developers #AI #Chips

Thank you to Caroline Hyde for speaking with our CEO, Rodrigo Liang as he discusses the path to better scaling and efficiency in AI on Bloomberg Technology. "AI inferencing is going to be 10x more of a footprint than training, so we're going to have to think about how to be energy-efficient, how to be cost efficient before everyone gets to have it." - Rodrigo Liang ?? Watch his full interview here ?? https://lnkd.in/gVbjTh_s

A little BTS in #NYC where Rodrigo Liang sat down with Judy Khan Shaw for the #NYSEFloorTalk series at the NYSE.

Visiting Bloomberg News today - Rodrigo Liang spoke with Caroline Hyde about #AI, inferencing, and reactions to #Nvidia earnings.

Fine-tuning open-source models like Llama 3 8B can outperform GPT-4o for specialized tasks, providing results that are ... ?? faster ?? cheaper ?? and more accurate Read more ?? https://lnkd.in/gTurn_RF #OpenSource #AI #MachineLearning Varun Badrinath Krishna

At a Bloomberg Intelligence event in #NYC, Rodrigo Liang met with Matthew Miller to chat about the future of #GenAI. Here's the TLDR: 1?? 2025 is the year for #AI inference 2?? Zero reason to use an NVIDIA chip for inference 3?? Power-hungry GPUs & how SambaNova’s RDU’s use less power Rodrigo Liang went on to say “If you look at the NVIDIA chips, I think there is zero reason today for you to use an NVIDIA chip for inference. Zero other than maybe you like the logo. Because there are alternatives in the market today where you can run 10 times faster at one tenth the power. And frankly, as a customer, nobody really wants to be completely dependent on a single entity. So, we give you an alternative to create more variety in your supply chain. … Now we’re not telling everyone we can replace everything that they do. They did a great job actually building this market over 15-20 years. … We’ve seen over history lots of examples where companies get locked into a single technology vendor. Once you’re trapped it’s not easy to get out. Now is a great time to look at alternatives to add to your ecosystem. … Who know’s what’s going to happen next year or the year after that. But today what you’re seeing with NVIDIA technology is that it’s getting more expensive, and more power-hungry, and it’s harder and harder to get.”

Hey, SC Conference Series attendees! ?? Get ready for Birds of a Feather, where you can hear a series of talks presenting advances in tools, technologies, & use cases along with a discussion driven by you. We will have two sessions: 1?? Democratizing AI accelerators for HPC applications 2?? NAIRR Pilot Program See you there! #SC24 #AI #NAIRR

Live at Slush, Our Chief Technologist & Co-Founder, Kunle Olukotun chats with CNBC about how #AI demand is only going to increase in 2025! ?? Get all the insights & watch the video here?? https://cnb.cx/3ZbP1bz

Our Chief Technologist & Co-Founder, Kunle Olukotun took to the stage at Slush, Europe’s premier startup conference, alongside: ??? Thomas Wolf - Hugging Face ??? Grace Isford - Lux Capital ??? Ravi Mattu - The New York Times During the panel, Kunle Olukotun shared a compelling vision for how #opensource #AI is leveling the playing field, revolutionizing infrastructure, and empowering startups to innovate efficiently. Key takeaways from Kunle’s insights: 1?? Efficiency is the New Frontier: Open-source models, paired with optimized hardware, enable companies to reduce costs while achieving high performance. Kunle highlighted how inference-heavy applications are shaping the landscape, demanding faster token generation and lower energy use. 2?? Owning the Model: Open source empowers businesses to fine-tune models on their own data and own the outcomes, reducing reliance on proprietary systems while maintaining data privacy. 3?? Agentic Applications: The future isn’t just about standalone AI models but intelligent orchestration layers that combine specialized models for more complex workflows. Kunle called this out as a major shift coming in 2025. Kunle underlined that open source isn’t just a technology, it’s a philosophy driving inclusivity, innovation, and collaboration in AI. ??