DL

[DL @rootstocker]

@PaulSkallas @bryan_johnson is he on to something?

[DL @rootstocker]

@noisyb0y1 But that’s like an 80k setup

[DL @rootstocker]

@level941 LSLGB005G | Grand Seiko

[DL @rootstocker]

@paul_k_0907 Why wouldn’t they just use the same insurance model as cybercabs when in unsupervised FSD. You don’t think world models will make any difference to FSD?

[DL @rootstocker]

@paul_k_0907 We have Y also HW4. Honestly speaking, I can’t be bothered to drive now. Looking forward to next year when unsupervised FSD rolls out.

[DL @rootstocker]

@paul_k_0907 Do you have a Tesla with FSD?

[DL @rootstocker]

@NoLimitGains Time to short?

[DL @rootstocker]

@__tinygrad__ Can you do a single or double pro 6000?

[DL @rootstocker]

@__tinygrad__ Why not a single pro 6000 box?

[DL @rootstocker]

@HansAmato Report

[DL @rootstocker]

@NVIDIAGeForce PRAGMATA RTX

[DL @rootstocker]

@NoLimitGains Singularity

[DL @rootstocker]

@TokyoDeepValue Share. Please?

[DL @rootstocker]

@0xSero Doesn’t he show egpu is slower?

[DL @rootstocker]

@_QueenMarni @ick_real Haha right on

[DL @rootstocker]

@__tinygrad__ Cant tell if it’s serious or not

[DL @rootstocker]

@DataChaz I got a 5090ti 16 gb vram running Qwen 3.5 - 9B through LM studio, I’m only hitting 50 token/s

[DL @rootstocker]

@chddaniel Shipper

[DL @rootstocker]

@MCV492 @blutomus @ChrisLaubAI @grok @grok or maybe the nvida GPU route?