Compute derivatives

With the difficulty of mining a block directly linked to AI compute, rising when more AI compute joins the network and going down when it leaves, the difficulty functions as a public index of AI compute [1].

The TensorCash protocol offers options and contracts-for-difference on this index, so anyone can take exposure to AI compute, even without running a GPU themselves [1].

For miners, this offers more certainty about their future income. A position that pays out when the difficulty rises makes up for the lower income from the same GPUs. With a contract-for-difference, the miner gives up part of the extra income if the difficulty falls instead, so its income stays stable in both directions [1].

When a contract is set up, collateral in $TSC is locked and both sides agree on the block at which it ends. The chain keeps time in blocks, so this block works as the contract's end date. Once the chain reaches it, the chain checks the difficulty itself and settles the contract in $TSC. There is no need for an oracle, the outside service that would otherwise report this data to the chain [1].

References

1. [1]AI Compute Derivatives whitepaper.tensorcash.org/whitepapers/AI_Compute_Derivatives_Whitepaper-modern.pdf