AI inference is migrating from the datacenter onto the device — AI PCs cross half of all PC sales in 2026 and edge-inference silicon is scaling fast — but the value is broad and contested, splitting between an architecture tax (Arm), platform leaders (Qualcomm, Apple) and efficiency specialists, with no single winner capturing the on-device shift.
📈 What changed: AI PCs crossed more than 50% of 2026 PC sales with Copilot+ standardising a 40+ TOPS NPU floor and Qualcomm's X2 Elite Extreme reaching 80 TOPS — turning on-device inference from…
| Indicator | Score | Reading |
|---|---|---|
| Maturity higher is better · high confidence | 82/100 | Mature, shipping silicon; multiple competing edge architectures. |
| Evidence Strength higher is better · high confidence | 66/100 | Market-size and adoption figures corroborated across research houses; no peer-reviewed tier. |
| Commercial Proximity higher is better · high confidence | 85/100 | Revenue at scale now; the AI-PC cycle is the proven near-term wedge. |
| Capital & Policy Support higher is better · high confidence | 69/100 | Modest direct funding; data-privacy and efficiency rules pull demand on-device. |
| Crowding Risk lower is better · medium confidence | 68/100 | Crowded in the mega-caps; specialists less owned but Arm richly valued. |
| Reflexivity Risk lower is better · medium confidence | 50/100 | Earnings-anchored leaders with some valuation getting ahead of delivery. |
Every indicator score is computed by the Northstar engine from analyst-set ordinal bands — never hand-written, never stored.
The public page is the summary (backed by 8 cited sources). Full briefing — company map, catalysts, native signals and sources — is inside the app.
Open in the appStart free →The Northstar Signal — what moved across all 350 themes and why, in your inbox. Free, no spam, one-click unsubscribe.