Editor’s word: I took a number of weeks off from the column — and logged round 21,000 air miles principally going to conferences but in addition seeing Woman Gaga with my spouse — however we’re again! With that: I’m within the behavior of bookmarking on LinkedIn, books, magazines, motion pictures, newspapers, and data, issues I feel are insightful and fascinating. What I’m not within the behavior of doing is ever revisiting these insightful, fascinating bits of commentary and doing something with them that may profit anybody aside from myself. This weekly column is an effort to right that.
At its current Snapdragon Summit in Maui, Qualcomm CEO Cristiano Amon laid out the corporate’s imaginative and prescient for the “ecosystem of you.” The concept is that your constellation of non-public units — smartphone, watch, earbuds, glasses and PC — seamlessly work collectively, operating multi-modal AI fashions utilizing highly-personalized context to ship agentic experiences. Given Qualcomm’s presence in all of those system classes, and its ongoing execution in opposition to a big diversification technique, the corporate appears effectively positioned to learn from AI on the far edge: you.
This client framing for edge AI can be relevant to the huge, variable world of trade — a key piece of the aforementioned diversification technique. Simply as you and I might profit from context-aware, distributed intelligence, so too would cameras, robotic arms, sensors and different units current wherever from mines and farms to warehouses and manufacturing facility flooring. Because it builds out the “ecosystem of you” for shoppers, Qualcomm is in parallel constructing out an ecosystem of issues for trade.
Past designing the systems-on-chip that allow AI processing on edge units, Qualcomm has been steadily bolstering its software program, providers and developer-facing choices to make sure hardware-based innovation interprets into real-world influence for finish customers. Over the previous 19 months, the corporate has struck offers to amass, or deepen its partnership with, Foundries.io, Edge Impulse, Advantech and, within the newest, Arduino. We’ll unpack the element however the theme is obvious: Qualcomm is extending its management in edge semiconductors right into a full-stack edge AI platform that fuses {hardware}, software program and group.
Three acquisitions, one platform
In March 2024, Foundries.io was acquired by the Qualcomm Innovation Middle. The concept was to strengthen Qualcomm’s place within the embedded computing trade. Foundries.io “supplies beneficial experience within the growth and deployment of open-source software program and helps to efficiently commercialize merchandise that incorporate a Linux distribution.” The corporate operates as an unbiased subsidiary. Qualcomm characterised the acquisition as meant “to increase our open-source experience and velocity product commercialization with Qualcomm Linux.” Foundries.io “abstracts the complexities of creating and updating Linux-based IoT and edge units.”
A yr later, Qualcomm acquired Edge Impulse, which focuses on “creating edge AI options and deploying them to edge units throughout the worldwide {hardware} ecosystem…[and] enabl[ing] builders and enterprise groups to streamline and speed up the method of constructing, delivering and optimizing embedded AI and machine studying options utilizing real-world knowledge.” The acquisition additionally introduced greater than 170,000 builders utilizing Edge Impulse’s approachable, low-code method to AI/ML for edge units into Qualcomm’s orbit.
Two months later at Computex in Taiwan, Qualcomm introduced a collaboration with Advantech “to advance AI-driven IoT functions…enabling deeper integration of [Qualcomm’s] cutting-edge applied sciences into Advantech’s edge computing and edge AI platforms — accelerating the deployment of clever options throughout industries.” The 2 firms stated they might “foster developer innovation” and “ship a developer-oriented instrument chain.”
And this week, Qualcomm introduced an settlement to amass open-source {hardware}/software program agency Arduino. Along with the acquisition, Qualcomm and Arduino, which is able to keep its unbiased model, introduced the brand new UNO Q, a “dual-brain board” that makes use of Qualcomm’s Dragonwing platform, and the Arduino App Lab, a “growth setting that unifies the Arduino journey throughout real-time OS, Linux, Python and AI flows — simplifying construct, take a look at and deployment.”
Qualcomm’s Nakul Duggal, group normal supervisor for automotive, industrial and embedded IoT, tied all of it collectively: “With our acquisitions of Foundries.io, Edge Impulse and now Arduino, we’re accelerating our imaginative and prescient to democratize entry to our modern AI and computing merchandise for the worldwide developer group. Arduino has constructed a vibrant world group of builders and creators. By combining their open-source ethos with Qualcomm Applied sciences’ portfolio of vanguard merchandise and applied sciences, we’re serving to allow tens of millions of builders to create clever options sooner and extra effectively — together with a path in the direction of world commercialization by leveraging the size of our ecosystem.”
A blueprint for AI in all places
Taken as a complete, these strikes kind a purposeful plan. Foundries.io serves as a foundational unified Linux base, Edge Impulse is the engine for AI growth, Advantech is a bridge to convey these capabilities to trade and Arduino provides attain to tens of millions of builders from hobbyists to professionals.
To a level, this can be a vertical integration play and it’s a little bit of a roll-up, however it’s extra formidable than that. Qualcomm is forging the connections between units, builders and knowledge in a manner that’s tuned for particular person edges. The massive image right here is that whereas the cloud is educating machines to suppose, the sting is educating them to grasp.
Returning to Qualcomm’s diversification technique, Amon earlier this yr mentioned the sting AI tailwind the corporate was driving, and the way it suits into the bigger evolution from offering SoCs to offering platforms. “Our superior connectivity, computing and edge AI applied sciences and product portfolio proceed to be extremely differentiated and more and more related to a broad vary of industries…We count on that whereas coaching will proceed within the cloud, inference will run more and more on units, making AI extra accessible, customizable and environment friendly. It will encourage the event of extra focused, purpose-oriented fashions and functions, which we anticipate will drive elevated adoption, and in flip, demand for Qualcomm platforms throughout a spread of units.”
Whether or not it’s an ecosystem of you or an ecosystem of issues, Qualcomm appears to grasp that intelligence belongs near context and knowledge, and that’s what the sting is all about.
