Live data from Hacker News

Run Kimi K3 using 29 GB of RAM at 0.50 tok/s

github.com

141–150 of 181 posts

Re: Run Kimi K3 using 29 GB of RAM at 0.50 tok/s

#142
post #81

Standard llama.cpp can mmap the gguf, so it'll stay on disk if it doesn't fit on memory, and the kernel page cache will ensure the hot parts ("resident trunk") stay resident. What's the benefit of a custom implementation at all?

Letting the kernel use SSD based swap space for something this big would be a good way to destroy its cumulative write endurance over a period of just a couple months. I would be very interested in seeing SMART self reported drive cumulative write and wear out stats if this was done for more than a short test. In my experience llama-server is better run with --no-mmap on things that will fit entirely into RAM. Though…

When you mmap a file read only, the OS fills it in on a page fault by reading. There's no writes.

Re: Run Kimi K3 using 29 GB of RAM at 0.50 tok/s

#143
28 Jul 2026 21:35:19 UTC

Running Kimi K3 on a M1 Max

https://github.com/gavamedia/deltafin

https://news.ycombinator.com/item?id=49090233

[ok]

29 Jul 2026 02:01:04 UTC

Kimi K3 running on any device locally

https://github.com/RightNow-AI/local-kimi

https://news.ycombinator.com/item?id=49092591

[dead]

29 Jul 2026 13:49:32 UTC

Show HN: Waste Run the full 2.78T-parameter Kimi K3 on a laptop

https://news.ycombinator.com/item?id=49097512

[flagged] [dead]

29 Jul 2026 14:38:35 UTC

Self-hosting Kimi K3: 20% more hardware cost, 20% better task resolution

https://aistack.imec-int.com/blog/gpu-self-hosting

https://news.ycombinator.com/item?id=49098130

[ok]

29 Jul 2026 14:57:14 UTC

Run Kimi K3 on a local computer

https://github.com/sqliteai/waste

https://news.ycombinator.com/item?id=49098395

[ok]

29 Jul 2026 15:41:50 UTC

Show HN: A new engine to run Kimi K3 on a laptop

https://news.ycombinator.com/item?id=49098966

[ok]

29 Jul 2026 19:54:13 UTC

Kimi K3 for local use (1.56TB > 594GB) compressed by Unsloth

https://huggingface.co/unsloth/Kimi-K3-GGUF

https://news.ycombinator.com/item?id=49102192

[ok]

29 Jul 2026 20:02:40 UTC

Kimi k3 now runs on one consumer GPU

https://twitter.com/Akashi203/status/2082555972380401852

https://news.ycombinator.com/item?id=49102291

[ok]

30 Jul 2026 06:12:55 UTC

Show HN: Run Full Kimi K3 with 29 GB of RAM

https://github.com/sqliteai/waste/

https://news.ycombinator.com/item?id=49106591

[ok]

30 Jul 2026 13:05:09 UTC

Kimi k3 run on RTX 5090

https://github.com/RightNow-AI/local-kimi

https://news.ycombinator.com/item?id=49109455

[ok]

30 Jul 2026 14:01:44 UTC

A new inference engine to run Kimi K3 2.78T parameter with 29GB of RAM

https://marcobambini.substack.com/p/the-waste-inference-engi...

https://news.ycombinator.com/item?id=49110183

[ok]

30 Jul 2026 16:52:58 UTC

Running Kimi K3 on a local computer

https://github.com/sqliteai/waste

https://news.ycombinator.com/item?id=49112587

[ok]

Re: Run Kimi K3 using 29 GB of RAM at 0.50 tok/s

#144

Earlier quoted context omitted.

It's not a cost if you self-consume your energy, even if you get paid for what you inject. You are just amortizing differently the cost of the PV installation. But what you get for the energy you sell is not real money, it's just money you MIGHT consume as energy in other periods.

I have no idea what you're trying to say. With PV, if you consume one additional kWh your bank balance will be 20c lower than it would otherwise have been. You can argue semantics about whether that is a "cost" all you want, but practically the end result is the same and any financial decisions about whether to consume that kWh should have an identical result to what you'd decide if you bought the energy in for the s…

they're saying you might not use the extra energy if your PV runs year round. If you continuously had a year round surplus, you wouldn't have a need to buy any energy, so the credits would be monopoly money.

Re: Run Kimi K3 using 29 GB of RAM at 0.50 tok/s

#145
post #29

Approximate calculation is putting the cost at ~$5 per million tokens (assuming 42W sustained, 20¢/kWh), and that's excluding hardware and other costs.

I got to the same conclusion another way. There's ~2.6million seconds a month and this is getting 0.5tok/s which is 1.3million tokens a month. Give some room for overhead and a reasonable rule of thumb; The cost to run the machine per month is the cost per million tokens.

Shit, beats the hell out of GitHub CoPilot charging $.01 per "credit." For Opus that's one credit for 400 tokens genned, so $5 converts to 200k tokens down, plus the cost of whatever context you sent up.

Re: Run Kimi K3 using 29 GB of RAM at 0.50 tok/s

#146

Earlier quoted context omitted.

I have no idea what you're trying to say. With PV, if you consume one additional kWh your bank balance will be 20c lower than it would otherwise have been. You can argue semantics about whether that is a "cost" all you want, but practically the end result is the same and any financial decisions about whether to consume that kWh should have an identical result to what you'd decide if you bought the energy in for the s…

they're saying you might not use the extra energy if your PV runs year round. If you continuously had a year round surplus, you wouldn't have a need to buy any energy, so the credits would be monopoly money.

But unless you're off-grid you can always use extra PV energy by selling it to the grid.

Re: Run Kimi K3 using 29 GB of RAM at 0.50 tok/s

#147

Idk about you guys but I'd find 0.5t/s useless. Even for long tasks. I'd rather just shell out the money to offload as much as possible to say 2x 4060ti 16gb with tensor parallelisation. Anything but that low token rate. This is the sort of thing I'd expect in 20 years for some cyberpunk esque "turtlebot" that thinks at 0.5t/s, is solar powered and performs some menial civic maintenance background task like cutting g…

I pay $20 for codex, use it daily for coding, and still haven't dipped below 50% for weekly usage.

I wouldn't even be able to afford buying 29 gigs of ram, or a new video card with hardware prices the way they are now.

Maybe if it was 2016-2018 prices, I'd think about it.

Re: Run Kimi K3 using 29 GB of RAM at 0.50 tok/s

#148

Earlier quoted context omitted.

wow. shamelessly stealing the name of one of the most robust and effective libraries ever, to push some ai bullshit. what an embarrassment

We are backed by the SQLite author, and we have the right to use the SQLite name. Please do your homework before writing such comments. We used an Elastic license for some projects to protect our work from being used in SASS without a prior agreement. WASTE is and will always be available with a very permissive license. I really don't understand why some people prefer to spread hate instead of just asking for clarifi…

Even so, sqliteai squandered cultivating its reputation by using a non-free license for various projects.

Re: Run Kimi K3 using 29 GB of RAM at 0.50 tok/s

#150

Earlier quoted context omitted.

they're saying you might not use the extra energy if your PV runs year round. If you continuously had a year round surplus, you wouldn't have a need to buy any energy, so the credits would be monopoly money.

But unless you're off-grid you can always use extra PV energy by selling it to the grid.

The grid doesn't give you money in these locations.

They give you credits for future consumption. You can only spend this on energy. If you always have a surplus, you will just rack up large amounts of credits.

Post reply on HN