10-12-2026, 07:00 PM
Ran a quick check on my box. On M3 Max with SGLang, I saw basically the same behavior. The surprise was that memory pressure showed up before throughput collapsed.
learning in public
|
rackmount vs tower for homelab inference
|
|
« Next Oldest | Next Newest »
|
| Messages In This Thread |
|
rackmount vs tower for homelab inference - by latentpeak22 - 10-12-2026, 05:21 PM
Re: rackmount vs tower for homelab inference - by rocmcrate26 - 10-12-2026, 06:24 PM
Re: rackmount vs tower for homelab inference - by cozygrid12 - 10-12-2026, 07:00 PM
Re: rackmount vs tower for homelab inference - by vramghost - 10-13-2026, 05:41 AM
Re: rackmount vs tower for homelab inference - by tensorport - 10-12-2026, 11:31 PM
Re: rackmount vs tower for homelab inference - by homelabhenry - 10-13-2026, 04:18 AM
Re: rackmount vs tower for homelab inference - by tensorport - 10-13-2026, 05:15 AM
Re: rackmount vs tower for homelab inference - by rocmriley - 10-13-2026, 07:51 AM
|
Users browsing this thread: 1 Guest(s)