Myron

I met this issue too. Is there any way to fix it?

Theoretically, dual mining is workable. However, it seems like we're using different drivers, libs, and languages to implement SGX. It's not easy to do so. Meanwhile, we both need CPU...

not easy to do it 😞

testing

[{"name":"Trie read benchmark(empty database (2,000,000 keys), db_type: RocksDb)","raw_average":8442179,"average":8364764},{"name":"Trie read benchmark(empty database (2,000,000 keys), db_type: ParityDb)","raw_average":34920,"average":34669}] nanosecond

crust@VM-0-16-debian:~$ ./node-bench-bin ::trie::write::empty -j [{"name":"Trie write benchmark(empty database (2,000,000 keys), db_type = RocksDb)","raw_average":6002305,"average":5912002},{"name":"Trie write benchmark(empty database (2,000,000 keys), db_type = ParityDb)","raw_average":127991,"average":125365}]

[{"name":"Trie write benchmark(large database (50,000,000 keys), db_type = RocksDb)","raw_average":1725934,"average":1342061},{"name":"Trie write benchmark(large database (50,000,000 keys), db_type = ParityDb)","raw_average":1602140,"average":1446221}]

with same prefix keys [{"name":"Trie read benchmark(large database (50,000,000 keys), db_type: RocksDb)","raw_average":491864,"average":455012},{"name":"Trie read benchmark(large database (50,000,000 keys), db_type: ParityDb)","raw_average":465603,"average":460628}]

random [{"name":"Trie read benchmark(large database (50,000,000 keys), db_type: RocksDb)","raw_average":722812,"average":717637},{"name":"Trie read benchmark(large database (50,000,000 keys), db_type: ParityDb)","raw_average":270572,"average":256024}]

with same prefix keys [{"name":"Trie read benchmark(empty database (2,000,000 keys), db_type: RocksDb)","raw_average":9746887,"average":9713109},{"name":"Trie read benchmark(empty database (2,000,000 keys), db_type: ParityDb)","raw_average":456027,"average":410957}]