Sorso View is the operator-facing publication on the AI compute economy. Anchored in the Sorso View Compute Index, our weekly benchmark across nineteen GPU cloud providers, and broadened to the structural and civilizational implications of the largest infrastructure buildout in modern history.
Inaugural Index publishes soon. The first Sorso View Compute Index publishes the week of our launch, with the full methodology and editorial commentary alongside it. Subscribers receive it in their inbox the morning it publishes.
Join the listShort editorial notes, four to five hundred words, before the market open. New rate-card disclosures, hyperscaler announcements, capacity events, financial filings — read in the time it takes to finish coffee. The publication you can rely on as a daily habit, not a weekly project.
Read the briefs →The Sorso View Compute Index measures median hourly rental rates across nineteen providers in five tiers. Hyperscalers, specialized neoclouds, aggregators, cost-positioned operators, and inference-implied. Every number sourced. Every methodology disclosed. The record of what the market charged, week by week.
Read the Index →Long-form pieces on the structural and civilizational implications of the AI compute economy. Capital flows, market structure, labor and economy, geopolitics. Written with the same rigor as the data, for readers trying to make sense of a moment few publications are covering with seriousness and range.
Read the essays →The market for AI compute is opaque, fragmented, and full of confidently wrong answers. Our founding article on what we are building, why we are building it now, and the four commitments that hold us accountable.
The hyperscalers are committing the GDP of a mid-sized economy to AI infrastructure this year alone. Where the money is going, what they are betting on, and what happens to the rest of us if the bet is wrong.
The first weekly benchmark. Median H100 SXM, H100 PCIe, H200, B200, and A100 hourly rates across the nineteen-provider panel. Full methodology. Full source list. Full honesty about what the numbers show and what they don't.
Every number we publish is measured. Every estimate is marked. Every analysis stakes a position. We publish the honest read — including when the market doesn't want to hear it.
Two institutional financial data feeds — Silicon Data and Ornn AI — distribute GPU pricing indices via Bloomberg Terminal at institutional pricing. They serve hedge funds, market makers, and quantitative trading desks. They do not produce editorial content. They are not built for the people running this market.
The Sorso View Compute Index occupies the operator position. Free for the audience that actually moves AI compute spend. Methodology disclosed in the same place as the number. Every data point traceable to a public source URL with timestamp. Editorial commentary in every issue.
Three indices, three distinct purposes — the institutional traders have theirs, and now the operators have ours.