Qualcomm

Is your vehicle’s technology keeping up with the horsepower under the hood? The modern car is rapidly becoming the most powerful, connected device you own. But turning a car into a truly intelligent agent takes a whole new level of digital horsepower. If you want to track how software, AI, and next-generation tech are reshaping the entire automotive industry, we've got the place for you. Check out the best destination for industry trends, deep dives, and real-time updates on the shift to software-defined mobility. Follow Qualcomm Automotive, our go-to hub for the future of connected vehicles.

This Week in #AI: 🔵 We introduced Qualcomm Intelligent Multimedia SDK 2.0, a unified toolkit for building edge AI and multimedia applications on Qualcomm Dragonwing platforms: https://bit.ly/4hW4rKX 🔵 Arduino’s new VENTUNO Q, powered by Qualcomm #Dragonwing IQ8, brings physical AI and on-device GenAI to developers, with the first pre-order batch selling out in under an hour: https://bit.ly/45WnCgn 🔵 Read more to see how the same code that runs on NVIDIA and AMD GPUs can now target Qualcomm AI accelerators with the same developer experience and best-in-class performance: https://bit.ly/4i1iwXr

True AI leadership isn’t built on promises—it’s built on deployment. Over the past decade, Qualcomm has transformed from an entry player into a full-stack powerhouse driving physical, compute, and everywhere AI. Today, our Snapdragon architecture powers over half a billion vehicles on the road, brings 90M+ digital cockpits to life, and delivers commercial silicon for the next generation of robotics. We are taking physical AI from a $30B market today to a projected $1T opportunity by 2035, with the proven scale to back it up.

Exciting launch day for Arduino VENTUNO Q, powered by Qualcomm Dragonwing IQ8. Opened for pre-order today and sold out within an hour. 

We're moving from an ecosystem of devices to an Ecosystem of You. Imagine starting a task on your Snapdragon powered PC, continuing on your smartphone, and staying informed through your smartwatch. As agentic #AI advances, intelligent agents can preserve context across devices, creating experiences designed around the individual rather than any single product. Qualcomm President and CEO Cristiano R. Amon breaks it down:

This Week in #AI: 🔵 Fast Company turned to our own Durga Malladi to paint the picture of an Agentic AI future: https://bit.ly/4wCWJsH 🔵 BMW iX3 and Snapdragon Ride Pilot deliver 5 stars under Euro NCAP’s tougher 2026 safety protocols: https://bit.ly/4xdc58i

Is it a lane change? Or just a little drift? The Qualcomm AI perception stack uses 360° perception to create a birds-eye-view (BEV) so Snapdragon Ride can track vehicles’ trajectories continuously in real-time rather than just a single point-in-time and safely delay a lane change until the other driver steadies.

Qualcomm and Modular, a Qualcomm company had an exciting day of announcements at ModCon. Here are a few highlights:   -Modular MAX now runs natively on five silicon accelerator platforms including our own Qualcomm Dragonfly accelerators (AI200, AI250, and AI300), Amazon Trainium (T2-T4), Google TPU (v6-v8) and GPUs from NVIDIA (B200) and AMD (MI355). This is now the first truly open, high-performance, full-stack platform that is commercially available with support for broad heterogenous AI hardware.  -The Mojo 1.0 release is the foundation of our inference stack and Modular is making this open source under an Apache 2.0 licensing framework. This sends an important message to the industry that we want an open and collaborative approach to AI stack development. In addition, we have announced a partnership with Microsoft where Mojo will run natively on Windows, building on existing Linux and Mac support. -Modular also announced Modular Cloud completing the full-stack platform – Mojo at the kernel layer, MAX for model execution, and Modular Cloud for production-scale cloud deployment. Modular Cloud is already processing trillions of tokens daily across production workloads. This provides real-world validation of the stack while creating a hardware-agnostic deployment layer designed to optimize inference across different accelerators. The result is portability with performance: customers can choose silicon based on price, power, and availability rather than being locked in by proprietary software. -Qualcomm is forming an alliance program around Modular. During Cristiano R. Amon’s keynote, Qualcomm announced its intention to create an industry alliance program by year-end 2026. This underscores our commitment that Modular remains an open and heterogenous ecosystem spanning GPUs, CPUs, NPUs and custom silicon accelerators. This means Modular becomes a unifying software layer that benefits the industry, where hardware players compete below the abstraction layer. Qualcomm now has a compelling inference stack that lowers the deployment barriers for our accelerators into the data center. Early industry feedback has been very positive - stay tuned for further announcements as we continue to build out the ecosystem. 

What allows AI agents to reason, plan, and act across complex workflows? The answer isn't just more processing power. As AI workloads become more autonomous, coordination across components becomes increasingly important. Discover how CPUs orchestrate workloads across GPUs, NPUs, and DSPs to enable the seamless handoffs and real-time decision-making that power agentic AI systems: