What happened
Anza, the Solana Labs spinout that maintains the Agave client, published SIMD-0675 on Wednesday, per CryptoBriefing's writeup dated September 30. The document proposes that Solana's leader schedule, the ordered list of validators assigned to produce blocks in each slot, factor in the geographic location of validators when it's generated each epoch. Instead of a schedule that can hop from a validator in Tokyo to one in Frankfurt to one in Ashburn within three consecutive slots, SIMD-0675 would cluster runs of slots inside a single region before crossing to the next.
The idea is straightforward physics. Blocks and votes travel over fiber, and light in glass is slower than most people assume. Trimming the number of trans-Pacific and trans-Atlantic hops per second removes latency that the protocol currently absorbs.
The proposal is at the SIMD discussion stage. It has not been scheduled, tested on mainnet, or put to any validator vote.
Why it matters
Solana's whole product thesis leans on low block times and cheap execution. Every millisecond shaved from block propagation is a millisecond that doesn't have to be padded into the slot budget or eaten by skipped blocks. Regional clustering is one of the few remaining levers that doesn't require touching consensus itself.
It's a scheduling change, not a hard fork of the state machine. That's why the design is interesting even to operators who dislike it. The tradeoff is exactly the one Solana has been arguing about since 2021.