Running Qwen3.8-Flash-Next NVFP4 at 184 tok/s on a single RTX 6000 Pro

Posted by matteohorvath 2 days ago

Counter28Comment6OpenOriginal

Comments

Comment by IronWolve 2 days ago

Been running versions of this on my pc, if you have 64 gigs ram, you can run it on a 12+ gig cards easily.

Comment by aappleby 1 day ago

On "just" a RTX 6000 Pro? You mean one of those $16,000 gpus?

Comment by saejox 2 days ago

it is very hard to run 150k token context with these kind of setups. even if you do, all of us are used to running multiple agents at once.

i hope ram prices will come down a little.

Comment by lowbloodsugar 2 days ago

Been running Qwen3.8-Flash-Next NVFP4 for weeks now. What's new about this?

Comment by catchnear4321 2 days ago

hn title is completely different from the article title, which takes clicking the hn title to discover.

> Our first model-kernel combos

Comment by cjbprime 1 day ago

They did a specific-hardware kernel optimization for the Pro 6000 (available on AWS as g7e).