−10.4%
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 locally for privacy, speed, and cost. This basket spreads across the device makers, the edge chips that enable on-device inference, the memory and storage that hungry local models need, and a small slice of the open-model ecosystem itself. It's spread across many names rather than betting on one, with a slice of cash kept aside to add on dips, and I'm treating it as a two-year view rather than a permanent holding.
What it owns
- These companies sell the devices and systems where people and businesses run AI inference locally, from consumer PCs to corporate laptops to on-premises servers. · AppleIt has built neural processors into its devices for years and continues expanding on-device AI inference across iPhones and Macs., NvidiaIts RTX consumer GPUs are the standard hardware for running open-source models on home computers and workstations., Dell TechnologiesIt sells corporate laptops and workstations that get refreshed when companies upgrade their fleets to run AI locally., HP Inc.It is one of the world's largest PC makers, so an AI-driven upgrade cycle directly affects its unit shipments and pricing., Super Micro ComputerIt builds compact servers that businesses deploy on-premises to run AI models locally instead of renting cloud capacity., Corsair GamingIt supplies memory, power supplies, and cases for consumers building custom PCs to run AI models at home.33%
- These chip designers and manufacturers build the processors and specialized AI accelerators that power local inference across phones, laptops, cameras, and industrial equipment. · QualcommIts Snapdragon processors include dedicated neural engines that power AI on Windows laptops and premium smartphones., Advanced Micro DevicesIt ships Ryzen AI processors with built-in neural processing units into AI-capable laptops, and consumer GPUs for local model inference., Arm HoldingsIts chip designs power the vast majority of smartphones and a growing share of laptops, all potential platforms for edge AI., BroadcomIt supplies the networking and connectivity silicon that edge devices depend on, plus custom chip design services for AI accelerators., Lattice SemiconductorIt makes small, power-efficient programmable processors that run AI workloads in cameras, embedded systems, and edge devices where efficiency matters most., AmbarellaIts processors run computer vision AI inside cameras and vehicles, processing video on the device rather than sending it to the cloud.27%
- Running AI models locally requires more memory and storage per device than previous generations, and this bucket captures the suppliers of those components. · Micron TechnologyLocal AI models require significantly more DRAM than previous PC generations, and memory pricing matters to the economics of the upgrade cycle., Western DigitalModel files and their training data live on local storage, and larger AI-capable devices push storage capacity requirements higher., Seagate TechnologyLocal AI systems generate and consume growing volumes of data that must be stored on drives at the edge and in nearby servers., SandiskIt manufactures the NAND flash storage inside SSDs that ship with AI-capable phones and computers.19%
- Open-source AI models and the foundries that manufacture all the chips in this take enable edge AI to function as a real alternative to cloud-only inference. · Meta PlatformsIts Llama open-source models form the foundation for capable AI running on consumer hardware rather than cloud services., Taiwan SemiconductorIt manufactures virtually every edge AI chip in this basket regardless of which designer wins market share.11%
- Cash10%