I started from zero with Forgejo, so Gitea is not a consideration. I just don’t need latest and greatest as long as security is covered. But I finally noticed that v15 is the new LTS, so that’s what I hope to run during the next couple of years.

Well, before your response, v11 was the latest LTS release, received security patches and did all I wanted it to do. Now, I actually did look at the new release and found out that 15 is a new LTS. So, thank you! I guess it’s time to review the breaking changes and get prepared.

I’m happily running v11 LTS - I have enough to do besides dealing with breaking changes every 3 months.

Wow, I didn’t think you were running 176GB worth of GPUs on a consumer board! I don’t have an extra board, and my gaming PC that has 9070XT is not a good basis for multi GPU build - it has a cheap mATX motherboard with too few slots and lanes. So it’s going to be a new build. Used EPYC boards look interesting for that.

Thanks! I was running some models on my RX 9070XT, but only Ollama works flawlessly. I couldn’t make llama.Cpp to run Gemma4 or the newer Qwen - maybe I’m hitting that incompatibility, but probably it’s the skill issue.

P40 doesn’t look very appealing. 32 GB V100 costs about the same as 2xP40, less VRAM in total, but it’s faster, will use less power.

But I’m not sure if I follow you on the PCIe… If I run a model that spans multiple GPUs, doesn’t PCIe bandwidth matter?

Thank you! This is really helpful. 32 GB V100 or pair of 5060ti s looks very interesting, and about the same price. Does running multiple GPUs require any special hardware? I mean apart from the motherboard with 2+ PCIe x16 slots?

Not this one. It IS a new shiny one, they coat $300-$400 more on eBay than in stores due to limited supply.

Nice! You kinda answered my next question already with this web tool. I was curious if you are getting any useful results from the model itself without feeding it with good data first or relying on hardcoded tools. 4b model must be really dumb for anything even little complicated. I see you recommend to run two models - is it in parallel or the router can control backend and switch models?

This looks awesome! Can you share the real life use cases for this? What are you using it for?

He’s been put in psychiatric institution today, reportedly

Does the increased density mean that the speed also goes up? It would be nice if a 7200 RPM drive could finally saturate SATA3 bandwidth.

All right, I think you are right on the curling upwards. I started another print, which is also using similar hexagonal sides, but is more massive - hexagons are larger, and walls are thicker. And it warps up visibly! When these overhanging arms on upper part of hexagons start printing, they just warp up while cooling. Then when nozzle comes back for the next layer, it hits the warped edges, pushing filament to the sides and making these artifacts. It didn’t break any segments, as they are thicker, but I bet that’s what happened in the original print.

On the new thicker print, the base of the model lifted up from the bed on one side at some point. I was watching the first two layers, so this happened some time later. Gonna try to slow down the print, lower nozzle temp a bit, bump up the bed temp a bit. Will see what happens.

But… but… but they are pretty!

Thanks! I’m a beginner, got the printer about two months back. I didn’t do any special tuning apart from auto leveling the bed.

I printed a temp tower with this filament, and it was okay from my viewpoint. Long overhangs were droopy on lower temperatures, and it had stringing from pointy ends on higher temps. I printed a few PETG prints before, but nothing with such fine details as this hexagonal thing.

Can you point me to some good source of info about tuning? For a noob like me it is hard to tell good resource from slop-compiled one…

Thanks again for the valuable feedback! I printed it again with higher temperature, 240. It didn’t change much, overall. Same print errors remained, however I got a few less of broken hexagons, so overall the print feels more sturdy.

Looks like the majority opinion here is to blame that all on wet filament. I don’t have the right equipment to dry it, so I’d probably just avoid printing similar parts with this spool, and take more precautions when it will be time for a next one. My drybox spool holder build is almost ready.

Thanks for another interesting point. It is faster than you are suggesting. My slicer settings for “external perimeters” is 25 mm/s and for “perimeters” is 40 mm/s. There are no other type of movements on these hexagonal structures.

I may try slower printing at some point, but it’s already a 7 hour print 😄

In the oven??? But wouldn’t it just melt? I have a gas oven, I don’t think it can go too low. But I will look it up. Maybe microwave is a way to go. Thanks for the tip.

That also makes sense. I imagine if there is a way to increase adhesion between layers that would help with such problem

2-3 weeks before I put it into a dry box with a bunch of silica gel

Thanks! I think I unpacked this filament about a month ago, and it was on open air for about two weeks, in 40-50% humidity. then I put it into a dry box with a box of silica gel, and it stays there at about 20% humidity. Until I take it out for a print. Is silica gel good enough for drying? I don’t have the drying machine.