@elonmusk @karpathy Can you prove that @Tesla HW3's 144 TOPS will be enough for FSD with the safety level necessary for regulatory approval and that @nvidia DRIVE Pegasus' 320 TOPS is excessive?
-
-
Because you are comparing against a 3 years old Drive PX 2. A chip that's about 3 generations old, without Tensor Cores for performing DL matrix operations, without HBM, without NVLink, without DLA, etc. 2016 vs. 2019, sounds fair. Why not compare with Turing chip architecture?
-
I’m currently using the GTX 2080 Ti and using CUDA for a compression app. The latency is still extremely high to the point where I can’t do real time compression.
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.