Launch tokens backed by real GPUs, paired with anything.
Pair your token with ETH, USDG, or a tokenized stock like TSLA. Every trade pays a fee, part of every fee buys GPU-hours, and anyone who has launched can spend them on any GPU on CudaPad — from a community DGX Spark to a B200 node.
- Compute raised
- $1.5K
- Available to spend
- $1.5K
- Launches
- 14
- GPUs backed
- 6
Every token funds a GPU.
Priced live from the curve. Filter by the class of GPU a token backs, sort by market cap, volume or progress.
All tokens
14 tokens
Common Questions
No. Fees go to the treasury, and the pool's share becomes GPU-hours that everyone who has launched can spend, including you. If you want fees paid to yourself, this is the wrong platform.
No. Launching once is what buys access, and it is permanent: your wallet stays eligible whatever happens to the token afterwards.
From GPU marketplaces. The keeper tops up an aggregator account (Shadeform-style: one API over RunPod, Lambda, Vast.ai and others) and jobs are provisioned there at the live rate. Desktop-class boxes like DGX Spark come from community hosts.
Yes. pons takes any ERC-20 as the pair asset, and on Robinhood Chain that includes tokenized stocks, ETFs and gold. The curve, the 2% fee and the graduation pool are then denominated in that asset; the keeper values claimed fees at spot before crediting the pool with its share.
Jobs are refused until the next claim refills it. Running jobs finish; nothing is queued and nothing is charged.
No. It is recorded at creation alongside the supply and the pricing, none of which can be changed afterwards.
No. There is no timelock that expires and no function that can move it.
We do. The keeper wallet funds its own transactions, and skips anything worth less than the gas it would cost to collect.
Compute-backed tokens.
Transparent settlement.
One transaction, and the model is funded. Fees are pooled, ledgered in micro-USD, and spendable by anyone who has launched.