−10.2%
since Sep 30
The idea
I believe AI is moving out of the cloud and onto the devices we already own, as open models shrink enough to run well on consumer chips. To express that, this basket spans device makers who sell the hardware, the edge silicon that powers on-device inference, the memory and storage that hungry local models need, and a small stake in who publishes the open models and fabricates the chips. It's a concentrated two-year view across four layers rather than single-name bets, with 10% held in cash to add on dips.
What it owns
- Companies that sell the end devices where AI inference happens locally, from phones to laptops to workstations to servers. · AppleIt has the largest installed base of devices with dedicated neural processors and has committed to running AI inference directly on iPhones and Macs., NvidiaIts RTX consumer GPUs are the standard hardware for running open-source AI models at home and in small offices., Dell TechnologiesIt sells the corporate laptops and workstations that get replaced when organizations upgrade their fleets for on-device AI., HP Inc.As one of the largest PC makers globally, it benefits directly from any refresh cycle driven by demand for AI-capable machines., Super Micro ComputerIt builds compact servers that businesses deploy on-premises to run AI locally instead of using cloud services., Corsair GamingIt sells memory, power supplies, and cases for custom PC builds used to run AI models at home, capturing the enthusiast segment.33%
- Chip designers and makers of processors, NPUs, and connectivity silicon that power on-device AI across phones, laptops, and embedded systems. · QualcommIts Snapdragon chips with dedicated AI processors are winning sockets in both premium phones and Windows-on-Arm laptops., Advanced Micro DevicesIt ships Ryzen AI processors with integrated NPUs into AI-capable laptops, and its consumer GPUs power local model inference., BroadcomIt makes the networking and connectivity silicon that edge devices depend on, plus custom chip designs for AI hardware makers., Arm HoldingsIts chip designs power most phones and a growing share of laptops, so wider AI device adoption means more architecture licensing revenue., Lattice SemiconductorIts low-power programmable chips handle AI tasks in cameras, laptops, and industrial devices where efficiency matters more than raw compute., AmbarellaIts processors run computer vision inference inside security cameras and automotive systems where processing happens on the device.27%
- Makers of DRAM, flash, and disk drives that store models and data on local devices instead of in the cloud. · Micron TechnologyAI PCs require substantially more DRAM than prior generations, and Micron supplies the memory that local inference demands., Western DigitalIt supplies the local storage drives that hold model files and the data that on-device AI systems process and create., SandiskIt supplies the NAND flash that goes into the SSDs in AI PCs and phones, benefiting from higher storage capacity per device., Seagate TechnologyIts drives store the expanding volume of data that local AI systems generate and consume across homes and edge locations.19%
- Companies that release or manufacture the open-source AI models and chips that make local inference practical and affordable. · Meta PlatformsIts Llama open-source models are the foundation that makes capable AI practical to run on consumer hardware and edge devices., Taiwan SemiconductorIt manufactures nearly every chip in this basket, from Apple's neural engines to Qualcomm's NPUs to AMD's processors.11%
- Cash10%