TechMakesArt
Region: US
Tuesday 31 March 2026 14:50:12 GMT
89004
3010
120
144
Music
Download
Comments
tiktokname :
I wish to have money to build a computer to run A.I. locally at home 🥺
2026-03-31 16:29:51
63
bbbk :
To train an AI like Gemini you need 30.000 H100 GPUs which costs 1 billion dollars. Then you need computers to install those h100s on and they should run for months. Overall it costs like 1.5 billion dollars to have your ow. AI.
2026-05-11 22:23:05
1
Xsals4 :
Just buy a Mac mini
2026-03-31 21:42:22
6
rrock :
5090 😀
2026-04-16 05:35:59
0
otterammo666 :
“whichllm” is a nice open source tool for finding models with quants for your specific hardware. really easy to use if you need a good starting point
2026-05-24 00:03:44
1
paul :
if you are gpu poor with a 3090, then what do you call me with my 1070ti?
2026-04-04 16:48:07
4
勇人 :
Would an RTX 5060 Ti 16GB be good enough? Or what would you recommend?
2026-04-09 16:38:34
0
Abdul | Web design/Cyber Sec :
So this means if I can run AI locally I dont need to pay for chatgpt anymore
2026-04-27 16:50:37
2
Aduktech :
What are the equipment needed to build a home lab
2026-04-05 21:45:16
0
astburysam :
I got hermes running with qwen3.5 9b, gtx 3060 works sweet! I also have another rig with a 1080ti 11gb using qwen3.5 9b, all works! Still trying to get Honcho (memory) setup but that is a bit tricky!
2026-04-01 10:04:28
5
Sandrinho4x4 :
I have a GTX1080 Ti with i think 11 GB. Whats best to run? I want it to crawl Websites and read them for lead gen
2026-03-31 18:56:10
2
mickr12770 :
why do I need great GPU if using as code or questions
2026-04-15 08:07:26
0
DragonDescendants :
sell all the hardware/copper and buy yourself a Mac mini :-)
2026-05-20 07:07:28
0
maxwell888 :
3070可以魔改16G三星s14显存，非常有性价比
2026-05-01 01:03:00
2
Just Facts, no BS. :
I used to mine... I have 12 RTX3060. Can I use them all in one system?
2026-04-13 07:25:09
1
MrJingles | Dating Chronicles :
Bazzite for the bc-250
2026-04-06 16:42:40
0
Robert Fontaine 🇨🇦 :
It's nice to be rich
2026-04-05 22:20:32
0
Omegote :
I have a 9070xt, I wonder what's the easiest path to start running local models with it
2026-03-31 19:41:12
0
user8547103211260 :
just release the videos already
2026-04-03 13:26:56
1
Nico Morgan | Equine Media :
How about running an LLM like Ollama on a machine which has integrated Intel graphics like a 12th Gen i5?
2026-04-01 10:05:23
3
AudioPhoenix :
hell yeah, been doing lots of local AI on my rx 7900 xt!
2026-05-08 00:50:08
1
redi_ville :
What the hell you do with the local AI?
2026-03-31 20:28:42
1
James Plotts Ek Fehu :
3090 w/24gb VRAM runs qwen3.5:35b nicely. powers my claw
2026-04-06 02:35:09
1
Andaluciasteve :
Compile Turboquant into Llama.cpp to make the most out of your Vram.
2026-04-01 00:43:11
1
LightHoneyLuo :
Follow
2026-03-31 21:14:02
2
To see more videos from user @techmakesart, please go to the Tikwm
homepage.
