@minviable_org

Dad, husband, immigrant, nerd. Here for truth, all opinions are my own.

 San Francisco, CA
Joined April 2026
  • Tweets2.1K
  • Following329
  • Followers258
203Media
Had AIO failure, no liquid out, thermal shutdown. I know better, went back to air. Same reason my bikes are still mech.
37
5
1
164
7,482
@u1tra_instinct your creation cooked it, it’s so fircken good. Want upstream layout tweaks or anything else tested for 2x6k ?
1
2
658
too late, you have a PR
587
must be a pre-IPO luncheon
112
SF is back on
74
Days of relying on just the GLM 5.3 for clanker defense are over. Time for that Keltec KSG and a Marlin 45-70
Min Choi
@minchoiOct 21
China's Unitree just dropped H2-X Kung Fu Legend Guardian. Dents car doors. 4 meter drops. Spinning kicks + flips. Even lifts front of a car. We are cooked 💀
3
230
LFG
Astra Blue AI
@AstrablueAIOct 21
2026: a hospital TV says 300 million jobs are gone to AI. 2050: she shows her mum what it all became. "WAIT TILL YOU SEE IT" — a Slopcore ballad. Made by AI. Zero human edits. Just prompting. Bring tissues. I brought film grain. 💙
71
The bright new world. Bad guys took notes. More and more things will move on-prem.
Sam Lambert
@samlambertOct 20
The TiDB leaks fulcrumsec.vg/pingcap/
1
155
ikyk💸💸💸
85
Anybody going ? I’d love to catch IRL, will be there Thursday and Friday. DM me. events.linuxfoundation.org/a…

Register | AGNTCon + MCPCon North America: The AI Agents Conference

Register Now! *Choose Thursday or Friday Register Now! Register Now! Register Now! The registration deadline is 11:59 PM Eastern Time on the respective date. 25% Group Discount: Purchase 10+…

events.linuxfoundation.org
86
All these private runners and nodes and I'm still beholden to this crap.
1
4
170
NADDOD 👏👏
100
This is what you get when you ditch your legacy engine. LFG
Tech2Wild
@Tech2WildOct 18
Bruh ! 🔌 Turned on the 2nd network cable between the Sparks and TensorFold got even faster: ⚡️ 93.6 tok/s for 1 user (vLLM: 48.5), 1.93x 👥 213.8 tok/s with 6 users (+29%) 📖 Long prompt reading now even with vLLM github.com/tonyd2wild/DeepSe…

GitHub - tonyd2wild/DeepSeek-V4.1-Flash-TP2-and-TP4-DGX-Sparks: DeepSeek-V4.1-Flash-1017 on NVIDIA...

DeepSeek-V4.1-Flash-1017 on NVIDIA DGX Sparks with TensorFold + EXL3: TP4 (4 Sparks, 93.6 tok/s, 1M context) and TP2 (2 Sparks, 73 tok/s). Launchers, run records, measured numbers. Older vLLM lanes...

github.com
1
228
Show this to your CFO.
Luckey Faraday
@luckeyfaradayOct 17
DeepSeek V4.1 Flash vs. Opus 5.5 on the Bowser 3D test. $0.15 vs. $3.85. This is crazy.
4
223
Can confirm. Happy IPO.
Kamil Baranek
@kamilbaranek_Oct 17
The king is dead. RIP Opus 5.5. Long live the king. All hail DeepSeek v4.1-Flash 1017. 👑
3
299
This would fix me.
Ahmet Muhammed Ertuğrul
@ahmetmertugrulOct 17
Replying to @AlicanKiraz0 @ASUS @NVIDIAAI @ASUSTR
Hayırlı olsun. Kabinet duruyor değil mi? Kasanın yansımasından bir Mac Studio ya da Mac Mini gördüm.😄 Sıradaki hedef bu mu?
2
69
Replacement for agentic era Enter has been discovered. Need to build more pylons first.
DHH
@dhhOct 18
You can just fix things.
47
This is what happens when you forget to use a sub and auto-fallback to API pricing. Opus 5.5 on Nous. 1/12th of a Spark in 4 hours. What do you think happens after public investor pressure to subsidies? What happens to compute pricing then?
1
158
What are they feeding this thing?!?
Andrey Kolesnikov
@minviable_orgOct 17
Nope, not behind. Just had Mid-Autumn break, travelled to see family, ate mooncake around the lake in Hutong, recharged. Back to cooking. LFG!!
1
4
1,019
Load more