Request
Add NexusPool to config/pools.json:
{
"name": "nexuspool",
"display_name": "NexusPool",
"host": "solo.nexuspool.io",
"port": 3350,
"operator": "NexusPool",
"pool_type": "solo",
"groups": ["all"],
"features": {
"sv2": true
}
}I checked the endpoint against the handshake collector/str_race.py performs: mining.subscribe with empty params, then mining.authorize [address, "x"]. It authorizes. The first mining.notify carries a non-empty merkle branch list with clean_jobs: true, so it should register under Full Template (winner_nonempty), not only Any Template.
solo.nexuspool.io is a latency-steered hostname across four regions: Los Angeles, Chicago, Frankfurt, and Singapore. Each vantage measures whichever region sits nearest to it. Stratum V2 runs on all four.
Why?
NexusPool is a registration-free, non-custodial solo pool at 0% fee. The miner's address goes straight into the coinbase: no account, no withdrawal.
I am not submitting this with a performance claim attached. We ran our own collector reproducing your Any Template vs Full Template distinction: 32 blocks per pool per vantage, from Los Angeles and Frankfurt, against pools already on your board, on 2026-09-11.
Any Template, ahead-or-tied on each shared block:
| vs |
LA |
Frankfurt |
| AtlasPool |
59% |
56% |
| parasite |
53% |
72% |
| eusolo_ckpool |
n/a |
81% |
| ckpool |
100% |
n/a |
| solo_cat |
100% |
n/a |
| braiins_solo |
n/a |
100% |
| solohash |
n/a |
100% |
| solomining |
n/a |
97% |
Median vs AtlasPool: 9.3 ms ahead in LA, 9.7 ms ahead in Frankfurt.
Full Template (winner_nonempty), ahead-or-tied on each shared block:
| vs |
LA |
Frankfurt |
| AtlasPool |
13% |
9% |
| parasite |
22% |
19% |
| ckpool |
100% |
n/a |
| solo_cat |
100% |
n/a |
| solohash |
n/a |
100% |
| solomining |
n/a |
94% (30 of 32) |
| braiins_solo |
n/a |
100% |
Median deficit vs AtlasPool: 107.8 ms in LA, 226.6 ms in Frankfurt. We trail AtlasPool and parasite on this race in both vantages. We would rather be measured on your terms, by your collector, whatever the result.
nexuspool.io/latency publishes a dated panel, as of 10 to 11 Sep 2026, for a different quantity than this board: time from our node seeing a new block's header to our first job being ready, 0.49 ms median over 162 mainnet blocks, ready before our own node finished validating the block on all 162, 205 ms ahead at the median. This is background context; the numbers above are the comparison.
TLS is not exposed on this endpoint; port 3443 is closed.
Request
Add NexusPool to
config/pools.json:{ "name": "nexuspool", "display_name": "NexusPool", "host": "solo.nexuspool.io", "port": 3350, "operator": "NexusPool", "pool_type": "solo", "groups": ["all"], "features": { "sv2": true } }I checked the endpoint against the handshake
collector/str_race.pyperforms:mining.subscribewith empty params, thenmining.authorize [address, "x"]. It authorizes. The firstmining.notifycarries a non-empty merkle branch list withclean_jobs: true, so it should register under Full Template (winner_nonempty), not only Any Template.solo.nexuspool.iois a latency-steered hostname across four regions: Los Angeles, Chicago, Frankfurt, and Singapore. Each vantage measures whichever region sits nearest to it. Stratum V2 runs on all four.Why?
NexusPool is a registration-free, non-custodial solo pool at 0% fee. The miner's address goes straight into the coinbase: no account, no withdrawal.
I am not submitting this with a performance claim attached. We ran our own collector reproducing your Any Template vs Full Template distinction: 32 blocks per pool per vantage, from Los Angeles and Frankfurt, against pools already on your board, on 2026-09-11.
Any Template, ahead-or-tied on each shared block:
Median vs AtlasPool: 9.3 ms ahead in LA, 9.7 ms ahead in Frankfurt.
Full Template (
winner_nonempty), ahead-or-tied on each shared block:Median deficit vs AtlasPool: 107.8 ms in LA, 226.6 ms in Frankfurt. We trail AtlasPool and parasite on this race in both vantages. We would rather be measured on your terms, by your collector, whatever the result.
nexuspool.io/latencypublishes a dated panel, as of 10 to 11 Sep 2026, for a different quantity than this board: time from our node seeing a new block's header to our first job being ready, 0.49 ms median over 162 mainnet blocks, ready before our own node finished validating the block on all 162, 205 ms ahead at the median. This is background context; the numbers above are the comparison.TLS is not exposed on this endpoint; port 3443 is closed.