Research
Company-authored research on independent evidence for AI compute — measurement, disclosure, GPU contracts, lender controls and financing. Each paper is published in full, with a fixed fingerprint, an independent timestamp and fully dated sources.
Dates and provenance
Each paper shows the date it states as its research cutoff. The file is published unaltered, with a checksum and an independent notarisation token attached for anyone who wants to verify it. Dates are evidenced by the public records each paper cites, not by file metadata.
Each paper carries a SHA-256 fingerprint, an RFC-3161 timestamp and an OpenTimestamps record.
Measurement and independent evidence
Why net connection-point behaviour under-determines both underlying load and the equipment producing it.
Research as at 7 May 20265 ppTimestamped
Evidence scope and data provenance
Scope, provenance and the limits of proportionate disclosure in computational-load evidence.
Research as at 7 May 20266 ppTimestamped
GPU contracts and capital
Four filed agreements separate technical proof, customer acceptance and funding.
Research as at 14 June 20267 ppTimestamped
Institutional workflow and automation
Where scoped independent evidence enters lender controls, and the three boundaries it cannot cross.
Research as at 1 August 20265 ppTimestamped
GPU finance and programme economics
Rental economics, capital structures and the evidence that can support a continuing deployment programme.
Current research 22 September 202611 ppTimestamped
Related
The methodology this research supports is protected by three provisional patent filings on the public register.