Stephen Foskett on Nostr: I mean it’s nice that the M3 E core is a lot faster but the way the macOS scheduler ...
I mean it’s nice that the M3 E core is a lot faster but the way the macOS scheduler works they’re not going to help much for interactive applications. In fact I bet the M3 Pro is slower than the M2 Pro in many interactive apps! What do you care if TimeMachine has two more cores to use when you’re compiling on the P cores? And does the scheduler treat those six E cores as one complex or two? Or three? Same for the P cores I guess. In all it’s a weird configuration.
Published at
2023-10-31 03:02:26Event JSON
{
"id": "6b5d04e549438246f1101b9be09b5d07356d5af54d293dddfe2faa113f9934b6",
"pubkey": "329cdad50ef1c817e516bb51338623e8d9e0dab25461b88aca8069f8ea33a54f",
"created_at": 1698721346,
"kind": 1,
"tags": [
[
"e",
"8480ac5be6e480f63a0b8f480553180b9139adf6e22bd26fefaefaa711500290",
"wss://relay.mostr.pub",
"reply"
],
[
"proxy",
"https://techfieldday.net/users/sfoskett/statuses/111327402154709452",
"activitypub"
]
],
"content": "I mean it’s nice that the M3 E core is a lot faster but the way the macOS scheduler works they’re not going to help much for interactive applications. In fact I bet the M3 Pro is slower than the M2 Pro in many interactive apps! What do you care if TimeMachine has two more cores to use when you’re compiling on the P cores? And does the scheduler treat those six E cores as one complex or two? Or three? Same for the P cores I guess. In all it’s a weird configuration.",
"sig": "ef7bb0a00a478654ae3b33c60896bfcf8b269e491cd1e87bd231f41b382b18e1696130b85b6c38883e2f21f23d2c89828bd5ed6c960216cdfb96d97db6635e8f"
}
