$TEL Telcoin Network Stage 2: One less round trip. 5,434 transactions per second. A validator goes offline. The chain keeps committing.
Picture two validators: one in London, the other in Seoul.
Thousands of miles apart, they must exchange messages to reach agreement on the blockchain. Every extra round trip adds time.
Stage 2 targeted one of those delays.
Previously, a vote request required an extra round trip to negotiate the communication protocol before the request could be sent.
The update changed that. The vote request now goes out with the protocol proposal, rather than waiting for a confirmation round trip first.
One less round trip before the vote request goes out.
In the initial latency re-check, the time from a validator’s own proposal to reaching its own quorum fell by roughly half:
London: 560 → 286 milliseconds.
Seoul: 264 → 138 milliseconds.
The Netherlands: 582 → 296 milliseconds.
The idle consensus round fell from 530 to 313 milliseconds.
Next came the 40-minute stress test. After the latency re-check, the minimum header delay was adjusted from 250 to 350 milliseconds to accommodate the slowest validator’s measured timing.
Ten validators. Ten regions: Los Angeles, Montréal, London, the Netherlands, Mumbai, Singapore, Hong Kong, Tokyo, Seoul and Sydney.
5,433.9 transactions per second average chain throughput, compared with 3,890 in the previous run — approximately 40% higher.
6,909.7 transactions per second during the best 60-second window.
2.44 seconds median time for a transaction to land, down from 3.38 seconds.
The generators submitted 14.36 million transactions. 99.7% landed.
Then came the outage test.
Continue here: https://x.com/angelofyhvh/status/2108714505064624369?s=46&t=Kxud7JVUlu_OLFIOPp14PA
https://benchmark.telcoin.network/runs/r20261007-0316/
Picture two validators: one in London, the other in Seoul.
Thousands of miles apart, they must exchange messages to reach agreement on the blockchain. Every extra round trip adds time.
Stage 2 targeted one of those delays.
Previously, a vote request required an extra round trip to negotiate the communication protocol before the request could be sent.
The update changed that. The vote request now goes out with the protocol proposal, rather than waiting for a confirmation round trip first.
One less round trip before the vote request goes out.
In the initial latency re-check, the time from a validator’s own proposal to reaching its own quorum fell by roughly half:
London: 560 → 286 milliseconds.
Seoul: 264 → 138 milliseconds.
The Netherlands: 582 → 296 milliseconds.
The idle consensus round fell from 530 to 313 milliseconds.
Next came the 40-minute stress test. After the latency re-check, the minimum header delay was adjusted from 250 to 350 milliseconds to accommodate the slowest validator’s measured timing.
Ten validators. Ten regions: Los Angeles, Montréal, London, the Netherlands, Mumbai, Singapore, Hong Kong, Tokyo, Seoul and Sydney.
5,433.9 transactions per second average chain throughput, compared with 3,890 in the previous run — approximately 40% higher.
6,909.7 transactions per second during the best 60-second window.
2.44 seconds median time for a transaction to land, down from 3.38 seconds.
The generators submitted 14.36 million transactions. 99.7% landed.
Then came the outage test.
Continue here: https://x.com/angelofyhvh/status/2108714505064624369?s=46&t=Kxud7JVUlu_OLFIOPp14PA
https://benchmark.telcoin.network/runs/r20261007-0316/