
AI infra capital partner. We originate, underwrite & manage financing for AI compute economy. Backed by @hack_vc, @FactionVC, @ambergroup_io, @AntalphaGlobal.
gaib.ai
Joined June 2024
- Tweets2.6K
- Following286
- Followers125K
- Likes2.7K
New episode of Where Capital Meets Compute: Wayne Nelms, co-founder and CTO of @OrnnExchange , on estimating GPU residual value when resale data is thin, using supply and demand models and what historical forward curves implied at the time.
New episode of Where Capital Meets Compute: @wayne_nelmz, co-founder and CTO of @OrnnExchange, on why financing a GPU still takes a five to seven year contract, and what a liquid compute market could change.
His argument: the biggest buyers of compute often have the least credit history, so lenders end up underwriting the asset itself. If years three through six can be hedged against a price index, a one or two year contract could be enough to finance the GPUs. In his words, "liquid compute markets will actually accelerate the pace of GPU deployment." We also get into why Ornn builds its indices from cleared transactions only, why he thinks the shortage is GPUs rather than data center space, and the comment Ornn filed with the CFTC arguing for compute futures on exchange.
This episode is sponsored by @ComputeDesk, which builds financial infrastructure for compute traders.
Their Compute Indexes publish daily reference prices for GPUs including the H100, H200, B200 and B300, and their platform helps teams price, procure, hedge, finance and settle compute from first offer to final agreement. Learn more at computedesk.com.
Tune in to hear Kony on Wednesday covering compute financing with @gate_ventures!
teapawt.org/i/spaces/1dKOlAeRdXOjJ…
Gate Ventures Podcast | Episode 21
High Rates, Heavy Iron: Who Gets Funded When Compute Costs More to Finance
As AI infrastructure spending accelerates, who can actually secure the capital to build and scale compute capacity?
Join Gate Ventures' Tiffany and Kony, Co-Founder & CEO of GAIB @konyk001 @gaib_ai , as we explore:
🔹 How AI data centers are financed through equity, debt, and prepayments
🔹 Why smaller operators struggle to secure funding and what makes GPU-backed loans viable
🔹 Whether tokens create real compute utility or simply reward holders
🔹 How onchain financing and compute hedging could unlock capital for neoclouds
🔹 The key signals that could reveal whether compute financing is working
📅 November 18 | 11:00 UTC | 19:00 GMT+8
📍 Video livestream on Gate Ventures' X — also live on X Spaces.
Open models let anyone serve near-frontier AI. The bottleneck is GPUs: labs and hyperscalers locked up new supply, and the same chip costs ~2x more at a hyperscaler than a GPU cloud.
Our take: lend against the GPU, not the renter.
Thanks @fluence_project for having us 👇
The same GPU costs about 2x more at a hyperscaler than at a GPU cloud. The four largest hyperscalers plan about $730B of capex in 2026.
Who’s winning the race for GPUs? We put prices, capex and supply side by side in The AI Compute Race Report, with @near_ai @gensynai @Compute_Labs @fortytwo @Theta_Network @PhalaNetwork @BoundlessHQ @gaib_ai @ionet 👇
GAIB retweeted
"There's people who will buy that." Andrawes Bahou of Compute Desk splits a compute deal into three risks: price, utilization and counterparty credit. He says each one has buyers with deep pockets.
Full episode link in bio.
New episode of Where Capital Meets Compute: Andrawes Bahou, Founder and CEO of Compute Desk, on splitting compute risk into price, utilization and counterparty credit, and the prop firms, asset disposition companies and traditional commodity traders starting to take each one.
New episode of Where Capital Meets Compute: @andrawesbahou, founder and CEO of Compute Desk, on why the off-take contract may be the wrong hedge for compute.
His argument: an off-take mainly protects against idle GPUs, but it also locks in price. Operators who signed three-year terms six months ago have since watched prices rise roughly 40%, with no way to reprice.
We get into phantom capacity ("the hopes of a cluster"), why Compute Desk scrapped its first index built on on-demand prices, and who is now willing to buy price, utilization and counterparty credit risk.
This episode is sponsored by @ComputeDesk, building financial infrastructure for compute traders.
Their Compute Indexes publish daily reference prices for GPUs including the H100, H200, B200 and B300, and their platform helps teams price, procure, hedge, finance and settle compute from first offer to final agreement. Learn more at computedesk.com.
2
11
1,818
Glad to be listed on @DIAdata_org 's Global Stablecoin Map!
.@gaib_ai's AID is on the Global Stablecoin Map.
Explore supply by chain, peg history, holders and much more for AID.
Link in the tweet below.
GAIB retweeted
The old GPU cloud pitch: great products on top, high margins on the clusters. Evan Conrad heard it as a buyer at an AI lab.
Even modeling out to "mega huge, giant company," they were pretty sure they'd never pay for those products. Margins were thin.
Full episode link in bio.
GAIB retweeted
SF Compute wants to make the GPU hour the collateral.
Evan Conrad on the loan he has in mind: backed by a piece of paper that says "I own this cluster from this time to that time," with a liquid market to sell it into if the customer falls through.
Full episode link in bio.
GAIB retweeted
A GPU hour is not a GPU hour. Evan Conrad of SF Compute compares it to a "car hour": it matters a lot what you're renting.
His take: cash settlement can get compute futures off the ground, but not without physically settled data. There isn't any yet.
Full episode link in bio.
GAIB retweeted
On-demand is "the teeniest, tiniest fraction" of the GPU market. Most of it is reserved, and the two don't move together.
Evan Conrad of SF Compute on why hedging one with the other is like hedging oil with power.
Full episode link in bio.
New episode of Where Capital Meets Compute:
@evanjconrad , Founder and CEO of @sfcompute, on making the GPU hour usable as loan collateral, and why that takes a liquid order book, physical settlement and an operator that owns no hardware.
New episode of Where Capital Meets Compute: Evan Conrad, founder and CEO of SF Compute, on why the company had to stop brokering compute and start running the whole hotel.
SF Compute began as a small AI lab about $5 million in the hole on a GPU cluster it had to sublease every month to survive. Today it operates clusters it doesn't own and sells the capacity on a physically settled forwards market where every buyer writes their own spec, with several hundred million dollars of volume so far ("not huge, but we're scaling").
We get into why he thinks credit is the real bottleneck for GPU clouds, and how SF Compute wants to make the GPU hour usable as collateral.
This episode is sponsored by @computedesk, which builds financial infrastructure for compute traders.
Their Compute Indexes publish daily reference prices for GPUs including the H100, H200, B200 and B300, and their platform helps teams price, procure, hedge, finance and settle compute from first offer to final agreement. Learn more at computedesk.com.
1
20
1,891
GAIB retweeted
There's a 19 gigawatt gap between AI data center demand and supply right now, widening toward 36. So are we late in the bubble, or not?
Thomas Galbraith, founder and CEO of Barkr, argues the question is framed wrong. A correction is coming, but it won't land evenly: hardest on companies selling revenue expectations they can't meet, lighter on the infrastructure underneath.
Full episode available now.
New episode of Where Capital Meets Compute: Thomas Galbraith, Founder and CEO of Barkr, on why his firm guarantees its own GPU valuations, and why it prices off resale transactions rather than rental data.
New episode of Where Capital Meets Compute: Thomas Galbraith, founder and CEO of Barkr, on what it takes to price hard assets accurately enough to stand behind the number.
Barkr has run about $40B in valuations this year, backs roughly $500M of them with a Munich Re-backed guarantee, and works only from resale transactions. We get into how its GPU depreciation view has changed, the difference between a residual value guarantee and residual value insurance, and why he expects the coming correction to land unevenly.
This episode is sponsored by @computedesk, which builds financial infrastructure for compute traders. Their Compute Indexes publish daily reference prices for GPUs including the H100, H200, B200 and B300, and their platform helps teams price, procure, hedge, finance and settle compute from first offer to final agreement. Learn more at computedesk.com.
1
11
2,229
Next: a $140M pipeline of qualified GPU deals ready to deploy.
Full Q3 breakdown on the blog 👇
blog.gaib.ai/said-q3-2026-10…
102






