Solana Foundation finds lower-stake validators lost larger vote-credit share at 250ms

A Solana Foundation study found that lower-stake validators lost a larger share of reward-linked vote credits at the 250-millisecond slot target than the stake-weighted network average. The credit-loss fraction was 1.6360% with validators counted equally, compared with 0.0874% when weighted by stake. The Foundation published the study on Sept. 28. Its September changelog listed the 250-millisecond feature gate as active on Mainnet. The study found that skip rates stayed low and broadly stable as slot targets shortened. Votes took more slots to land, with the largest latency increases among nodes in Asia and South America. Network-average vote latency remained well below two slots. The Foundation said it saw no evidence of consensus instability. It said small samples of regional leader handoffs could not rule out a statistical fluke in apparent regional skip patterns. Vote credits help determine staking rewards, but neither credit-loss fraction represents a reward-loss percentage. The study did not quantify individual SOL payouts or isolate shorter slots as the cause of the gap. The Foundation treats 200 milliseconds as a possible next target, not an active Mainnet setting. The Foundation says its planned Alpenglow design would change how validators send votes, limiting comparisons with future voting performance.