DePIN / Compute.

Crypto rails for the GPU shortage — until the shortage ends, then we find out what is structural.

Tracked 03Conviction 02Watching 01Skeptical 00

Thesis

Why this sector, why now.

Decentralized compute caught the AI capacity-crunch spillover. That is a real wedge today and a real question for 2027. The thesis is that aggregating consumer + small-data-center GPUs creates a permanent cost advantage versus hyperscalers. The counter-thesis is that hyperscalers catch up on capacity and decentralized supply only wins on the long tail (privacy, geography, censorship resistance). I track both. Pick projects with real institutional customers — not Discord-bot rendering jobs.

Signals I track

What would move my read.

01
Average GPU utilization rates across networks (proxy for real demand)
02
Institutional case studies — which AI labs train on which networks
03
Tokenomics surviving low-utilization periods

Kill shot

What would kill the thesis

Hyperscaler GPU prices fall 30%+ and decentralized supply becomes uncompetitive on cost-per-flop without privacy / geographic constraints.

Coverage

Projects on the radar.

io.net

Conviction

Decentralized GPU network — aggregator model — institutional traction

Deep read →

Render Network

Conviction

Veteran decentralized rendering — pivoted to AI compute — Solana migration done

Deep read →

Akash Network

Watching

Cosmos-based decentralized cloud — earliest DePIN cloud — slower growth

Deep read →

POV

Long-form in this sector.

DePIN / ComputeMay 18, 2026· 2 min

DePIN Compute: io.net, Render, and the Structural Question

Both networks have real GPU supply and real customers. The structural question — whether decentralized compute is a meaningful price-improvement over hyperscaler rentals at scale — has not been answered. Until it is, neither network has an unassailable moat.

Read piece →

Going deeper

Bespoke DePIN / Compute dive for your fund.

Brief me

Draft thesis · Grey voice in progress, edits land continuously