5
votes
$\begingroup$

Before I start, I want to let you know that I am completely new to the field of deep learning! Since I need a new graphics card either way (gaming you know) I am thinking about buying the GTX 1060 with 6GB or the 1070 ti with 8GB. Because I am not rich, basically I am a pretty poor student ;), I don't want to waste my money. I don't need deep learning for my studies I just want to dive into this topic because of personal interest. What I want to say is that I can wait a little bit longer and don't need the results as quickly as possible.

Can I do deep learning with the 1060 (6GB seem to be very limiting, according to some websites) or the 1070 ti? Is the 1070 ti overkill for a personal hobby deep learner?

Or should I wait for the new generation Nvidia graphics card?

$\endgroup$
1

3 Answers 3

7
votes
$\begingroup$

Regarding specific choices I can't recommend, but if you are completely new, you should probably learn/code some more until you get a GPU. There is a lot to learn in machine learning before GPU speedups make a significant difference, and until then doing the computations on any old CPU would be just fine, especially if you are just starting since you won't be doing anything too complex. You will know when computational resources are your main bottleneck, and until then it shouldn't really matter too much.

Or, you could also rent computing power from say, AWS or Google

$\endgroup$
5
votes
$\begingroup$

Given that you're a student doing this out of personal interest and wanting to do some gaming on the side, I'd suggest the GTX 1060 6GB since at present the GTX 1070Ti is overpriced due to crypto miners (this will date the answer, but for reference the 1060 is going for ~GBP340, the 1070Ti for ~GBP600; two other options are the 1050Ti 4GB for ~GBP160 or the vanilla 1080 at ~GBP650).

'Which GPU...' by Tim Dettmers is very helpful, as is 'Picking a GPU...' by Slav Ivanov, especially the summaries at the end for different use cases. As you're not looking at spending a huge amount of money, the 1060 seems like a good compromise as the 1050Ti might just leave you with a disappointing gaming experience. Finding a used 1070 is also suggested, but you'd need to be comfortable with that.

Other answers have mentioned the cloud, but that doesn't help with your gaming. If you want to save some cash while you're waiting for the next gen of cards, take advantage of your student status on AWS educate or Azure on MS Imagine - the GitHub student dev pack is a good package.

$\endgroup$
3
votes
$\begingroup$

I don't think you need to invest in any kind of GPU unless you're familiar with the computations required for the task you want to achieve using deep learning.

Also, by the time you've sufficiently mastered Deep Learning to a point where you can actually make the most of your GPU, there will be new products in the market.

So until then I suggest you use your CPU for doing little tasks such as Regression etc. You can always use the free credit offered by the various cloud companies for your tasks

$\endgroup$
1
  • 1
    $\begingroup$ You make a good point about holding off on hardware purchases b/c newer, more powerful products continuously emerge, which has the effect of reducing the price of previous products. The longer you wait = more bang for your buck! $\endgroup$
    – DukeZhou
    Jan 26, 2018 at 21:29

Not the answer you're looking for? Browse other questions tagged .