atultw

Could you share some of those applications which are better solved with an FPGA? As a student I have some ideas but am interested to hear more.

Exactly, I know because I did the same thing!

If you capture a video and SLAM map of the whole space, you could use some VQA model like cosmos reason offline to extract key points and descriptions. Maybe even plan the route offline for the open ended task like “clean kitchen”. Then load the route and all you need is localization and obstacle avoidance

Admittedly, my use of CUDA and Metal is fairly surface-level. But I have had great success using LLMs to convert whole gaussian splatting CUDA codebases to Metal. It's not ideal for maintainability and not 1:1, but if CUDA was a moat for NVIDIA, I believe LLMs have dealt a blow to it.

New term for web scraping just dropped

To do gaussian splatting anywhere near in real time, you need good depth data to initialize the gaussian positions. This can of course come from monocular depth but then you are back to monocular depth vs lidar.

Nice work

Kalman filters so real. I think part of it is that LLMs suggest it as a solution all the time

“A Drop of Freedom” reminds me of Apple's old flower wallpapers with the black background

minimap exists in Xcode but it's not quite the same. It only shows `// MARK: ` comments.

Would love for Xcode to have more language support. Maybe an lsp client.