ZK Proving Costs Soar as Trillion-Dollar AI Data Centers Compete for GPUs

The zero-knowledge (ZK) proving market is facing a new challenge: competing for the same Graphics Processing Units (GPUs) as trillion-dollar AI data centers. Cysic CEO Leo Fan told crypto.news that GPU use has become a binding constraint for ZK proving because proof systems now compete with heavily funded AI data centers for the same silicon.

Fan explained that AI models are converging, but compute isn't. 'Compute isn’t. Everyone assumed proving costs would fall because chips get cheaper. Instead, we’re bidding against trillion-dollar data centre budgets for the same silicon,' he said.

The pressure on ZK proving costs is not just due to a lack of computing capacity, but also an architectural mismatch between zkVM software and the accelerators used to generate proofs. This leaves part of the available GPU capacity unused and raises the cost of each proof.

Cysic's Venus proving engine has exposed this mismatch by reducing the time spent coordinating work between CPUs and GPUs. The company recorded a 9% performance gain without replacing the underlying hardware, showing that existing accelerators were not being fully used.