I have some CUDA code that I need to run as fast as possible, so if I was going to blow $1500, my use case would imply that I should go with two 3080s.
However, I also play around with deep learning stuff, expect to do so more in the future, but don't currently follow it so closely.
Would someone care to ponder on what difference they think a 24GB gpu vs a 10GB gpu will have as a tool for deep learning dev over the next 3 years?
For what it's worth, I'm a computer vision guy, but I did have a play with DeepSpeech earlier this year.
However, I also play around with deep learning stuff, expect to do so more in the future, but don't currently follow it so closely.
Would someone care to ponder on what difference they think a 24GB gpu vs a 10GB gpu will have as a tool for deep learning dev over the next 3 years?
For what it's worth, I'm a computer vision guy, but I did have a play with DeepSpeech earlier this year.