Tue, 21 Apregulation

Comment P2P.org a créé un expéditeur de transactions Solana pour les équipes qui ne peuvent pas se permettre de manquer un créneau

Burns Brief

Solana traite plus de 162 millions de transactions quotidiennement à des heures de créneau d'une durée moyenne de 390 millisecondes. Les acteurs du marché évaluent soigneusement les implications, le résultat dépendant probablement des conditions macroéconomiques et du volume plus larges. Surveillez la réaction de $SOL : un mouvement décisif au-dessus ou en dessous des niveaux clés confirmera la prochaine tendance.

Solana processes over 162 million transactions daily at slot times averaging 390 milliseconds. For most users, that speed is more than sufficient. For trading firms, arbitrage bots, and liquidation engines, it is barely enough margin to work with. The difference between landing a transaction in slot 0 and landing it in slot 2 is not a rounding error. It is the difference between a profitable execution and a missed opportunity with fees already paid. On Solana, landing late is not free. Priority fees paid to win a slot are still charged when the transaction arrives after the opportunity is gone. This is the problem that P2P.org built Syncro Sender to solve. The real bottleneck is not Solana. It is the path to the leader. Most teams submitting transactions to Solana are using public RPC endpoints. These are designed for accessibility and general use, not for execution-critical workflows. They share bandwidth across thousands of concurrent users, offer no prioritization for time-sensitive transactions, and route through a constrained set of paths with no guarantee of directness or delivery speed. Research found that Stake-Weighted Quality of Service is the most effective mechanism for reducing transaction landing latency across all transaction types, outperforming both priority fees and Jito tips. Standard public RPC endpoints, those not peered with a staked validator, cannot access SWQoS priority bandwidth. They compete for the remaining approximately 20% of leader capacity alongside every other unstaked connection on the network. The result is structural: teams relying on public RPC are competing for the remaining 20% of available bandwidth, regardless of how much they pay in priority fees. Fees influence ordering after a transaction arrives. They do nothing to improve the probability that it arrives at all. This is not an API problem. It is a network design problem. How Solana transaction routing determines execution outcomes. What makes Syncro Sender different from

Key Takeaways