3 ms·
Google used reinforcement learning to design next-gen AI accelerator chips
- dorkmind 5y agoI was pleasantly surprised to learn that the wiki page for CUDA has a list of open problems in CS research: Accelerated rendering of 3D graphics Accelerated interconversion of video file formats Accelerated encryption, decryption and compression Bioinformatics, e.g. NGS DNA sequencing BarraCUDA[55] Distributed calculations, such as predicting the native conformation of proteins Medical analysis simulations, for example virtual reality based on CT and MRI scan images Physical simulations, in particular in fluid dynamics Neural network training in machine learning problems Face recognition Distributed computing Molecular dynamics Mining cryptocurrencies BOINC SETI@home Structure from motion (SfM) software I'm sure what Google is doing is higher up the food chain, but I think this list is a good starting place for undergraduates, especially juniors, who want to focus on CS research. I would only add bio-statistics specifically to this list and mention the possibility of applying AI to bio-engineering. Do you think this would appeal to post-secondary students?