prefill bottlenecks on longer prompts on A6000

Thread Rating:

0 Vote(s) - 0 Average
1
2
3
4
5

Linear Mode

prefill bottlenecks on longer prompts on A6000

byteyard Offline

Offline

Junior Member

Posts: 31
Threads: 5
Joined: Sep 2026
Reputation: 0

#10

10-11-2026, 09:41 PM

I tested this tonight. On GB200 with Ollama, I saw basically the same behavior. Turns out 'just one more test' was a lie.

Find

Reply

« Next Oldest | Next Newest »

Messages In This Thread

prefill bottlenecks on longer prompts on A6000 - by batchscope72 - 10-08-2026, 08:26 PM

Re: prefill bottlenecks on longer prompts on A6000 - by tensorworks - 10-08-2026, 08:41 PM

Re: prefill bottlenecks on longer prompts on A6000 - by routearch - 10-10-2026, 12:33 PM

Re: prefill bottlenecks on longer prompts on A6000 - by deepseeked - 10-08-2026, 09:34 PM

Re: prefill bottlenecks on longer prompts on A6000 - by tinycontext - 10-08-2026, 10:16 PM

Re: prefill bottlenecks on longer prompts on A6000 - by modrunner27 - 10-08-2026, 11:10 PM

Re: prefill bottlenecks on longer prompts on A6000 - by turboscope - 10-09-2026, 12:24 AM

Re: prefill bottlenecks on longer prompts on A6000 - by homelabhenry - 10-11-2026, 12:35 PM

Re: prefill bottlenecks on longer prompts on A6000 - by benchplease - 10-11-2026, 09:22 PM

Re: prefill bottlenecks on longer prompts on A6000 - by byteyard - 10-11-2026, 09:41 PM

Re: prefill bottlenecks on longer prompts on A6000 - by batchpeak - 10-13-2026, 10:18 AM

View a Printable Version

Forum Jump:

Users browsing this thread: 1 Guest(s)