@techno.karen: Two guys decided to launch local models of ai. Did it lead to something? Everything is in that video #techtokusa #aipc #macmini

Techno Karen🇺🇸
Techno Karen🇺🇸
Open In TikTok:
Region: US
Tuesday 11 August 2026 20:02:55 GMT
114605
7420
106
206

Music

Download

Comments

xeno.ku
Bro :
so whats the solution bro💔
2026-08-11 20:25:33
229
fortniteuser140
FortniteUser140 :
Cheap and ram don’t go in the same sentence
2026-08-12 00:11:51
473
doniczkaskibidi
doniczka :
isint 1000w like 0.1 cent
2026-08-17 22:19:27
7
oscarthemilllionaire
FinnKenealyT_T :
600 watts isn't even that bad
2026-08-12 15:04:16
61
diezgradoscentigrados10c
45 °C :
0.5 B is devil speaking 😭😭😭🙏
2026-08-18 00:12:45
1
best.horalka
˙☦︎⊹✝︎Horalka✝︎⊹☦︎˙☤☣︎ :
W patrick
2026-08-13 21:56:43
20
bgpushy
/Users/pushy :
I mean, Bonsai 27B (MLX variant) exists if you want reasonable inference on small devices like the Mac Mini (the model files weigh only 5.2GB). This exact model also exists in the GGUF format. The hardware in the second PC (dual Xeon CPUs and an Nvidia Tesla P40) could have its power limits cut (heavily on the CPUs) or be undervolted to not consume that much power, could run the GGUF version of Qwen 3.6 35B A3B with MTP at Q4_K_M or Q4_K_XL, GPT OSS 20B at Q8 (or FP8, both are similar), or Qwen 3.8 27B at Q6_K, all at reasonable speeds (all models fully fit in the P40's 24GB of VRAM)
2026-08-16 17:31:16
2
ilovefurries02
I love furries :
AI uses VRAM. not ram
2026-08-12 00:42:14
35
lukethehappyboy
Luke 🇬🇪🇪🇺 :
w spongebob
2026-08-11 20:13:33
6
kitatheranter
KitaTheRanter :
btw your airfryer pulls 1.5K Wats
2026-08-12 08:11:02
1
dml.bedo
CanB :
now let's see the consumption of both 💀💀
2026-08-12 08:51:35
9
εζ🥉 ≈
zepon :
wait a minute the Tesla Nvidia GPU has no display output 💔 it's for a server rack
2026-08-11 21:15:45
4
mamojestemspidermanem
皮塔饼 :
The Tesla p40 is really slow and has old architecture
2026-08-13 09:25:36
3
spirix_and_iorny
Spirix And Irony :
You gotta do a switch 1 vs switch 2!
2026-08-11 20:07:57
5
that.one.bbi.guy
the forgotten from tboi :
why would need it?
2026-08-12 00:42:01
1
dialordxs
DialordXS :
16gb ram struggling to run 0.5B model? wth bro and half of trillion parameter model in 24gb+24 vram? hmm what an dream
2026-08-12 14:56:18
1
gdenlatam
GDEnLatam :
I dont know if anyone else noticed, but Patrick (which is the dumbest normally) is MORE Intelligent that Spongebob in these videos!
2026-08-11 22:48:52
0
andreierbnescu
nuno :
128 gb of ram even if is old is gona cost more than 300 even for ram
2026-08-11 20:27:25
5
adriiuu11
Adriiuu🫧🤍 :
Spongebob is dumb ASF why is he buying the 16GB one for LLMS…😭
2026-08-16 08:27:58
4
sirjerryoperesionofficia
Sir Jerry O. Peresion :
Thinkcentre?
2026-08-12 09:46:40
2
francesco.cer
fuxk.fra :
me just knowing what are ping and fps
2026-08-16 20:57:56
1
lukas_beau1
LukasB :
yo
2026-08-11 20:05:36
2
baked.potato888
baked_potato888 :
hi
2026-08-11 20:06:47
1
zeroedits44148_2
Dr.[REDACTED] :
Isn’t that you like 2000 bucks
2026-08-17 16:42:37
0
To see more videos from user @techno.karen, please go to the Tikwm homepage.

Other Videos


About