Went to sleep at 3:30a last night trying to track down a nasty CUDA illegal memory access bug, woke up early to keep digging, & just now pushed a fix. Hard to beat the endorphin rush of doing battle with a tricky bug and coming out on top.

Jun 26, 2022 · 9:06 PM UTC

29
22
7
949
Load newest
Replying to @gdb
Congrats on finding it! I miss the CUDA days 😢 multiplexing variables to same stack space to engage all cores, setting up calls to be warp-able… are you running checksums after every few steps? How do did you debug it? Did CUDA me check ever start being easier to use?
1
1
Replying to @gdb
That's why we do it! Congratulations
Replying to @gdb
I have same problem can you help me
Replying to @gdb
Also can we use memory safe languages for CUDA
Replying to @gdb
Pushing a first fix and tweeting it is very bad for your mental health. You tweet it AFTER it's all done.
Replying to @gdb
You’re thought process here is admirable but mistaken. It’s not you vs a bug It’s you vs you and you vs whoever laid out the code