@SignOfficial #SignDigitalSovereignInfra #sign $SIGN
There is a point in the way SIGN describes Arma BFT finality that deserves far more scrutiny than it usually receives.
In the whitepaper, wCBDC wholesale operations are compared to traditional Real-Time Gross Settlement systems, with the claim that wCBDC offers “RTGS-level transparency.” It also states that Arma BFT delivers “immediate finality upon block commitment.” Both systems are therefore presented as having immediate finality, almost as though they are functionally equivalent.
They are not.
The reason each system can achieve immediate finality is fundamentally different, and that distinction matters.
Traditional RTGS finality is immediate because the system is centralized. A single institution, the central bank, maintains the authoritative ledger. Once the central bank records a settlement, that settlement is final because there is no competing version of truth. The speed comes from centralization. There is no consensus process to complete because there is only one authority.
Arma BFT finality works differently. It is a Byzantine Fault Tolerant consensus system, which means finality is only achieved when enough nodes in the consensus set agree on a block before it is committed. The whitepaper states that Arma can tolerate up to one-third Byzantine nodes, meaning that even if a meaningful portion of participants are faulty or malicious, the correct outcome can still be reached. Here, speed does not come from centralization. It comes from the efficiency of the BFT protocol.
But BFT consensus carries a basic liveness requirement. It depends on enough honest nodes being online and able to communicate for the system to continue operating. If too many nodes go offline — not because they are malicious, but simply because they are unavailable — the network can stop progressing. Traditional RTGS does not face this same constraint. A central bank can continue processing settlements even when other participants are offline. Arma BFT cannot move forward once the honest-node threshold falls below the level required for consensus.
For national CBDC infrastructure, this distinction is not minor.
A network partition, where nodes can communicate within separate groups but not across the full network, can divide a BFT consensus system into subsets that are unable to reach consensus. RTGS does not partition in that way. It may slow down when a participant is unavailable, but the central authority continues processing settlements.
The whitepaper frames Arma BFT’s immediate finality as an advance over traditional systems. In terms of throughput and decentralization, that is true. But when it comes to availability under network partition and node-failure conditions, the tradeoffs are different from RTGS, and those differences are not explained.
That leaves the core question unresolved: whether Arma BFT’s immediate finality is truly equivalent to RTGS finality for national payment infrastructure, or whether it is instead a stronger guarantee under normal operating conditions that comes with different failure modes in adversarial or degraded network environments.
