10-17-2026, 11:33 AM
Tried to reproduce it just now. My result was mixed: first run looked bad, second run looked normal. The surprise was that memory pressure showed up before throughput collapsed.
|
PCIe lane weirdness with inference loads
|
|
10-17-2026, 11:33 AM
Tried to reproduce it just now. My result was mixed: first run looked bad, second run looked normal. The surprise was that memory pressure showed up before throughput collapsed.
10-17-2026, 12:43 PM
Quote:Tried to reproduce it just now. My result was mixed: first run looked bad, second run looked normal. The surprise was that memory pressure showed up b Need a little more detail here. Do you see the same thing with a smaller batch or shorter prompt?
Mostly CUDA, sometimes wrong, usually measured
|
|
« Next Oldest | Next Newest »
|