Hacker News new | past | comments | ask | show | jobs | submit

Running Gemma 4 26B at 5 tokens/sec on a 13-year-old Xeon with no GPU

https://www.neomindlabs.com/2026/06/08/running-gemma-4-26b-at-5-tokens-sec-on-a-13-year-old-xeon-with-no-gpu/
I have a prediction. By the mid of 2027, we will have >200B MoE models running on basic consumer hardware.

I am running Qwen3.6-35B-A3B locally on my 16GB mac with 7-9 tokens/second. Link - https://github.com/deepanwadhwa/samosa-chat

This is a GPT4 level model running locally with a decent speed on a 16gb ram macbook air.

loading story #48924897
loading story #48924924
loading story #48926180
> on a decent speed

But you said 7-9 tokens/second, that's not a decent speed. I'm not an expert by all means but in my local experiments, less than 12 to 16 tps is too slow.

I think that any workflow that requires the user to stare at the tokens being generated live is using it wrong. Delegate, don't stare!

https://mikeveerman.github.io/tokenspeed/?rate=10&mode=text

You think of an idea that you want to have the LLM process, queue it up, and go back to what you were doing. Once you've finished reading the next article on HN about a 5 tps Xeon, your task will be complete. It's kind of like using a 3D printer: It doesn't matter if a print takes 10 hours, because when you come back in the morning it will be done.

Yes, with top-tier GPU farms you can hit hundreds of tokens per second. But if the old Xeon in the closet can get useful work done at 5 tokens per second, there are lots of people and lots of use cases where a free, unlimited 5 TPS stream is worth more than paying a dollars per day to get access to a 500 TPS source.

loading story #48925173
loading story #48925680
We clearly have different goals. I want an LLM to review my code, not the other way around.
loading story #48925098
loading story #48924813
loading story #48924753
loading story #48924928
loading story #48924752
loading story #48924658
loading story #48924678
loading story #48924725
loading story #48925527
loading story #48924540
loading story #48924821
loading story #48924641
loading story #48924779
loading story #48924791
loading story #48923723
loading story #48926078
loading story #48923267
loading story #48925857
loading story #48922524
loading story #48924373
loading story #48924594
loading story #48925472
loading story #48925283
loading story #48924959
loading story #48924168
loading story #48924332
loading story #48922789
loading story #48923829
loading story #48924376
loading story #48924760
loading story #48922435
loading story #48924517
loading story #48924770
loading story #48923851
Unfortunately, the post comes off as AI-written. Why not just write your own posts?
Not everyone is as confident in their writing as they are in their engineering