Any tips on figuring out why linkXCFramework times...

# kotlin-native

[Peter Hsu]

Any tips on figuring out why linkXCFramework times between different machines vary so much? e.g. on one project, an M1 Pro/32GB of RAM takes about 2 minutes per architecture. On an M2 Max/32GB of RAM takes about 4 minutes per architecture. On our CI machines which are Studio Ultras with 60GB of RAM, it similarly takes close to 4 minutes per architecture. Also any tips on improving these link times?