RTX2060 v RTX2070

[youngy84]

Hi, does anyone have an opinion on the 2060 v the 2070?

I’ve been reading slot online and many are saying the 2070 is not worth the extra money, the FPS and usage being marginally better. The main diffference bar the cuda cores is the 6 gig in the 2060 v the 8 gig in the 2070.

The laptop I’m buying (recoil RTX 17”) doesn’t gave a 4K screen so would a 2060 do the job?

Thanks!

 

[Oussebon]

Either is great. Remember that the RTX 2060 and 2070 in laptops aren't as powerful as their desktop counterparts, and also unlike desktop PCs you cannot upgrade thew graphics card. So there is an argument to be made for the 2070 over the 2060 in a laptop, if the price difference isn't too brutal.

If money is an issue though, the 2060 is a very solid pairing for 1080p gaming.

 

[youngy84]

Either is great. Remember that the RTX 2060 and 2070 in laptops aren't as powerful as their desktop counterparts, and also unlike desktop PCs you cannot upgrade thew graphics card. So there is an argument to be made for the 2070 over the 2060 in a laptop, if the price difference isn't too brutal.

If money is an issue though, the 2060 is a very solid pairing for 1080p gaming.

Ah, so all the reviews I would have been watching/reading would be based on a desktop one? I had no idea they were different... (console gamer, pc newbie) Is that what the Max q is? I guess I can find the specs online can't I?

Money is not really an issue, got about £2k and you have already suggested a build im likely to go with (Recoil RTX and it had the 2070 in it) I just got reading and its been slated quite a bit, maybe best going for the 2070 though if the laptop ones are less powerful. (I wondered how it would fit in the chassis duh)

 

[Oussebon]

In the past, you had desktop GPUs e.g. GTX 970 and laptop GPUs, which were differentiated with the letter M after them e.g. GTX 970M. Laptop GPUs were much, much weaker than their desktop counterparts (mostly due to power efficiency and there being only so much heat/power draw laptops can handle, as I understand it).

Last generation, Nvidia dropped use of the M to differentiate between desktop and mobile GPUs. This is arguably quite confusing for consumers, although in fairness to Nvidia the mobile GPUs had become much closer to their desktop counterparts in performance, though still lag behind.

Nvidia also introduced Max-Q variants of their mobile GPUs. These are focused on trying to find a balance between thermals and performance specifically to go into thinner/smaller chassis. It usually involves lower clock speeds so as to lower power usage, and so lower temps, but also lower performance.

There are reviews specifically of laptop GPUs. Notebookcheck.net has a table of benchmarks to give a rough idea of performance in games of various GPUs. By default the table only shows laptop ones but can also be made to show desktop ones.
https://www.notebookcheck.net/Computer-Games-on-Laptop-Graphics-Cards.13849.0.html
This is a truly excellent source of information, but it's (inherently) not perfect. Laptop chassis and desktops usually have different CPUs, so an entirely 'fair' test isn't possible. Also you can control the variables when benchmarking a desktop GPU - you can test multiple different GPUs while keeping all other components the same, and you can control airflow. With laptops, you can't really separate the GPU from the chassis. Each chassis may have a different cooling solution for example, which will affect the boost frequencies the laptop's graphics card can sustain.

But it gives you a rough idea

Also note that nearly all laptop GPUs are soldered onto the motherboard, as is the CPU. The motherboard is the laptop. Unlike a desktop where you can freely remove the GPU and upgrade it. you can't do that with laptops. Some laptops have removable GPUs in MXM form, but even then there's no guarantee future gens will be compatible.

 

[youngy84]

Oussebon,

Thank you for the replies, very grateful for you taking the time to reply and help out. I think based on that I will stick with the 2070.

Thank you.