Kinara, Inc.

🚀 Exciting news from SiFive and Kinara! We've teamed up to launch a USB-based enablement board that offers bare-metal access to the SiFive X280 RISC-V vector processor. Powered by Kinara's Ara-2 with dual X280 cores and high-efficiency NPUs, this board supports up to 40 TOPS for edge AI inference and generative AI workloads like LLaVA, LLaMA, and other LLMs. Now developers can prototype and evaluate X280 IP in real time and explore cutting-edge AI performance at the edge. Read the full press release: https://lnkd.in/gPvcFebt #RISCv #EdgeAI #GenerativeAI #LLMs #SiFive #Kinara #AIinference

Thrilled to host Satya Nadella our booth at TiEcon! We showcased the power of Kinara, Inc. Ara-2 processor, running Llama 3 8B and LLaVa, as well as Lenovo Neo Ultra platform—which integrates Ara-2 to enable real-time transcription, summarization, and translation at the edge. We had a great conversation covering Azure AI Foundry, BitNet, and the future of Agentic AI. Thank you, Satya Nadella, for stopping by and engaging with our work—your visit was a true highlight!

Thank you for your invaluable support, Kinara, Inc., in sponsoring #TiEcon2025, where we explore AI trends, tech, and the evolving industry landscape. Kinara AI delivers powerful Edge AI processors designed for Generative AI and vision-based applications—bringing intelligence closer to where data is created. We appreciate your role in advancing edge computing and shaping the future of AI innovation. #KinaraAI #EdgeAI #GenerativeAI #VisionAI #AIProcessors #AIInnovation #RealTimeAI #TechPartnership #TiEcon2025 #FutureOfAI

In partnership with #NXP, we built a demo based on LLaVA, a multimodal Large Language-and-Vision Assistant that can understand video, images, and text. Why is this interesting? Is there another low-cost Edge AI system that has been able to do this? 🔍 The impressive demo has already generated much interest from corporations wanting to explore the great potential of this application. See it for yourself in this video https://lnkd.in/gSjmVUG4 or in person (NXP will have an enhanced version of this in its Embedded World booth).

In partnership with #NXP, we built a demo based on LLaVA, a multimodal Large Language-and-Vision Assistant that can understand video, images, and text. Why is this interesting? Is there another low-cost Edge AI system that has been able to do this? 🔍 The impressive demo has already generated much interest from corporations wanting to explore the great potential of this application. See it for yourself in this video https://lnkd.in/gSjmVUG4 or in person (NXP will have an enhanced version of this in its Embedded World booth).

Today, NXP announced that they have entered into a definitive agreement to acquire Kinara. This acquisition will enhance and strengthen NXP’s ability to provide complete and scalable AI platforms, from TinyML to the multi-modal generative AI capability of Kinara’s Edge AI accelerators. Following this acquisition, the companies will create tighter integration of solutions to deliver scalable AI platforms for a variety of industrial and automotive AI inference needs.    We are thrilled for this next milestone in Kinara’s history and we offer a big thank you to NXP for this opportunity. And a big thanks to our employees and investors for the unwavering belief in Kinara and its technology.  https://lnkd.in/gDyJiaf7 #kinaraai #kinara #edgeai #nxp #ai #genai #deepseek

https://lnkd.in/gSCbfHxN

We've got quite an amazing demo lineup this year at #CES2025. For the first time on an edge AI processor, try out an enormous LLaVa-7B model running on our Ara-2. LLaVA, a leading-edge Large Language/Vision Assistant model, understands images in great detail and reports back in natural language. We'll also be demoing a) the generation of amazing images in near real-time with Stable Diffusion from complex text prompts; b) simultaneous real-time object detection with 4-8 cameras using a state-of-the-art YOLO model. Set up your personalized meeting with Kinara CEO, Ravi Annavajjhala, in our hospitality suite at the Venetian Hotel on January 7, 8, or 9. You'll have the opportunity to try out these fully-functional demos first hand. You have to see it to believe it! Contact us at info@kinara.ai