Hacker News Reader

If they are CPU throttled why not advertise egpu or igpu as requirement?

I’m more curious if the results are as consistently good as advertised, as usually NN results are cherry picked for good ones.

Odd. Works relatively well on my 2014 MBP. (With LR running in the background, too)

I mean, it goes further to your point -try it out. It seems that it depends _very_ much on the use cases.

I mean, if it's using DNN then you're going to be pretty much shit out of luck for perf unless your machine has an nvidia graphics card right?

If you’re using pretrained models then they should run ok. It’s the same reason why Tensorflow runs on phones: to run models, not necessarily to train models. DNNs take a lot of time to train.

TF runs at seconds-per-frame speeds on phones without dedicated accelerators.

Even inference is going to be slow when working on a 20 megapixel image. That's as much work as a 1220 sample batch at 128x128 resolution.