The four-category capital scorecard expanded with named transactions, burn-to-revenue context, and a per-category trend across recent issues. Each row corresponds to a row in the summary table on the issue page.
Category
Frontier Labs.
OpenAI, Anthropic, Google DeepMind, xAI
Capital in
~$95B→
vs ~$95B
Revenue out
~$21B→
vs ~$21B
Burn / rev
~1.3x
Lower means more capital out than in.
The read
Both S-1s remain at confidential-draft stage, with Deutsche Bank framing each offering at ~$60B. The strategic tell is pricing: closed labs are cutting rates ahead of the listings rather than harvesting margin. Investors should read the price cuts as demand-generation for the capacity buildout — and as pressure on every model-layer competitor's unit economics going into IPO season.
Trend across recent issues
Capital inRevenue out/$B per issue
Category
Hyperscaler-Hosted.
Azure-OpenAI, AWS-Anthropic, Google Cloud-Gemini, Oracle-OCI
Capital in
~$187B→
vs ~$187B
Revenue out
~$62B→
vs ~$62B
Burn / rev
~3.0x
Lower means more capital out than in.
The read
Aggregate 2026 top-4 capex of ~$725B (+77% YoY) is now colliding with operating cash flow, and the Epoch crossover analysis became the consensus frame this week. Meta selling raw GPU capacity and SoftBank standing up a 10GW-target neocloud are best read as responses to that pressure. Buyers gain negotiating leverage; investors should watch Alphabet's Jul 28 print as the first test of the crossover narrative against actuals.
The category absorbed two opposing signals in one week: fresh capital at rising marks (Together AI, Crusoe talks) and a violent public-market repricing of customer-concentration risk once Meta's compute-rental plan surfaced — Nebius carries a ~$27B Meta deal and CoreWeave ~$21B. Operators should diversify anchor-customer exposure in any multi-year neocloud contract; investors should now price hyperscaler competition into the category's cost of capital.
This week’s transactions
2026-07-01
Together AI Series C at $8.3B valuation, led by Aramco Ventures, plus >500MW of investor-capitalized compute commitments
Enterprise GPU clusters, sovereign and national programs, Cisco / Dell / HPE
Capital in
~$94B→
vs ~$94B
Revenue out
~$36B→
vs ~$36B
Burn / rev
~2.6x
Lower means more capital out than in.
The read
A reported ~$590B Korean government-plus-chaebol chip-hub plan needs primary confirmation before it moves this category. For hybrid buyers the real news is procedural: the Jul 9 FERC intervenor deadline and PJM's July filings will determine whether large-load interconnection accelerates or becomes a regional litigation map. Site strategy should still start with power and grid process, not GPU SKUs.