Hi team,
I'm a finance PhD researcher (Bocconi University) building an event study on GPU pricing around the DeepSeek-V3 release. The hypothesis is that frontier H100 SXM5 rental prices compressed more than commodity-tier GPU prices after the shock — a clean test of the within-vendor commoditization channel.
I need historical MsgCreateBid and MsgCreateLease transactions, with the matched GPU attributes (vendor, model, RAM, interface) and pricing (uakt / uusd), for the block range ~19,800,000 to ~20,500,000 (roughly Jan 20 – Feb 10, 2025).
Public RPCs have pruned this range — the deepest free archive I could find only goes back to ~Feb 2026. Since console-api already parses bid manifests into relational tables via the bid-screening / gpu-attribute-parser pipeline, would someone with read access to the production Postgres be willing to run a one-off SQL export to CSV for that window?
I will prominently cite Akash Network and the maintainers in the paper, and I'll share the final clean panel back with the community.
Thank you!
Rubén Fernández-Fuertes
PhD candidate, Finance — Bocconi University
fernandezfuertesruben@gmail.com
Hi team,
I'm a finance PhD researcher (Bocconi University) building an event study on GPU pricing around the DeepSeek-V3 release. The hypothesis is that frontier H100 SXM5 rental prices compressed more than commodity-tier GPU prices after the shock — a clean test of the within-vendor commoditization channel.
I need historical
MsgCreateBidandMsgCreateLeasetransactions, with the matched GPU attributes (vendor, model, RAM, interface) and pricing (uakt / uusd), for the block range ~19,800,000 to ~20,500,000 (roughly Jan 20 – Feb 10, 2025).Public RPCs have pruned this range — the deepest free archive I could find only goes back to ~Feb 2026. Since
console-apialready parses bid manifests into relational tables via thebid-screening/gpu-attribute-parserpipeline, would someone with read access to the production Postgres be willing to run a one-off SQL export to CSV for that window?I will prominently cite Akash Network and the maintainers in the paper, and I'll share the final clean panel back with the community.
Thank you!
Rubén Fernández-Fuertes
PhD candidate, Finance — Bocconi University
fernandezfuertesruben@gmail.com