Stripe's co-founder just stated that blockchain may need 1 billion TPS to support an AI agent-driven future.

This is where the significance of this number lies—and the current state of the industry.

Patrick and John Collison wrote in their annual letter for 2025 that AI agents may soon handle most internet transactions, while the current blockchain infrastructure is far from ready.

They provided a real example: on a major blockchain, the surge of memecoins resulted in payment delays of over 12 hours, and transaction fees skyrocketed by 35 times.

The current situation shown by Chainspect:

→ Solana: ~1,140 TPS (theoretical maximum: 65K)

→ ICP: ~1,196 TPS (theoretical maximum: ~210K)

→ Most networks: throughput of about 1,000 TPS

This is several orders of magnitude less than 1 billion.

Now consider the following scenario: in April 2025, when Qubic launches on its official L1 mainnet, transaction volumes reached 15.52 million TPS, a figure independently verified and published by CertiK. No Rollup, no L2 dependencies, zero fees, instant effect.

This is not a testnet simulation but an actual mainnet stress test, generating 1.518 billion transactions, verified by one of the most respected auditing firms in the industry.

Just 15.52 million TPS is still not enough to bridge the gap to 1 billion TPS, but it is the closest any active network has come to 1 billion TPS so far, and the gap is quite large. The underlying architecture (based on a tick consensus mechanism, with atomic execution and finality) is specifically designed to meet the high capacity and real-time computational demands required for AI workloads.

The discussion about infrastructure is shifting. Stripe has just told us what the future standards should be. The question now is, which networks are working towards this goal. #Qubic #TPS突破