Hot topic · builder-side · non-market

Making high-throughput actually usable

The conversation this week keeps circling back to the same practical question: once you can process transactions at massive scale, how do the tools and data structures keep up?

Teranode’s architecture (subtrees, microservices, unbounded blocks) is no longer just a theoretical claim. The quieter work is now about the pieces that sit underneath and around it.

On the storage side, Aerospike is being highlighted for delivering sub-millisecond UTXO lookups even under heavy load — the kind of speed required when you’re checking ownership at millions of transactions per second. At the same time, some builders are asking whether a more Bitcoin-specific UTXO store (modular, swappable, purpose-built) will eventually be needed as real throughput rises.

On the application side, LocalBSV is dealing with a real-world friction point: a minor UTXO selection bug that can interrupt BSV-21 token purchase chaining when a seller has a pending transaction. It’s not a fundamental design flaw, but it shows how wallet and indexer behaviour still need to be tightened for smooth multi-transaction flows. A fix is targeted before early October.

Taken together, the current pulse isn’t about announcing new capacity numbers. It’s about the quieter engineering required so that scale doesn’t just exist on paper — it actually works for the people building and using the network.

Sep 24, 2026