Post #1843648

I am totally melting this strix halo. I'm using this tool to run a memory system locally. I can even send the local model to do stuff as a subagent. And then someday, completely local. That makes sense to me, this is just the AI version of "hybrid cloud". The other workload maxing the CPU cores is building #gnomeos and #bluefin dakota, which I do all locally via this setup - test NUC reboots. https://www.youtube.com/shorts/cApfuO4WG9g Hey @kyle greg k-h was explaining his setup to me and now I am kind of motivated to see how people home lab this stuff out. I am going for the "locally driven testlab to make Project Bluiefin" because that would make a good talk lol. https://github.com/Siim/superlocalmemory