Upgrade or not?

Talk about Hardware used for Deep Learning


Locked
User avatar
MIK27
Posts: 1
Joined: Tue Jun 07, 2022 12:34 pm

Upgrade or not?

Post by MIK27 »

Hi,

Im impressed with faceswap and I like to boost my performance in this SW.

I have:
GIGABYTE Z390 M GAMING - Intel Z390
Intel Core i5-9600K
Patriot Viper VPN100, M.2 - 1TB
Seagate BarraCuda, 3,5" - 4TB
HyperX Predator 16GB (2x8GB) DDR4 3333
Seasonic Focus Plus Gold - 750W
GIGABYTE GeForce RTX 2070 SUPER GAMING OC 8G, 8GB GDDR6

I just bought pieces for upgrade:
3090ti with 24GB RAM
I also have Kingston FURY Renegade Black 32GB (2x 16GB) DDR4 4266MHz / CL19 / DIMM / 1.35V / Non-ECC / Un-Registered / 1Gx8
and i9 9900K

And now Im thinking about to sell my NIB GPU and buy 3090ti and maybe RAMs too and buy 2x32GB or even 2x64GB RAM..

Do you think it is worth it?

And I read somthing about enough high speed pcie slots and bottlenecking. My MOBO has just PCIe 16x 3.0 - is this enought? Maybe this is my botleneck in training..

Best regard
MIK

User avatar
torzdf
Posts: 2651
Joined: Fri Jul 12, 2019 12:53 am
Answers: 159
Has thanked: 129 times
Been thanked: 622 times

Re: Upgrade or not?

Post by torzdf »

3.0 should be fine, although I'm by no means a hardware expert.

Only you can really say whether the upgrade is worth the expense or not. If you are looking at doing multi-gpu, then it is sensible to match the cards, as you will be bottlenecked by the slower card.

My word is final

Locked