Add another 780 for SLI or go for 980 with the intention of SLIing in the future?

[JoeyMartin1958]

I currently own a GTX 780 3GB superclocked and am considering purchasing an additional GTX 780 and SLIing or selling my 780 and buying a 980 with the intention of then adding another 980 in a couple years (Or even the 980ti if you think it would be worth it in the long run). My goal for my computer is to run current and near-future games at high settings around 60fps on the Oculus Rift. Although we do not have the official specs of the Rift, it is rumored that the panels will be 1440p, but they could be slightly higher. There is also Windows 10 and Directx 12 to consider. The 980 supports all Directx 12 features and will benefit fully from its release, while older cards, such as my 780, will only receive some of the features and benefits. If I were to SLI 780s,I do not want to have to sell them both in 2 or so years just to then buy the newest graphics card. I do understand that future-proofing isn't real but I would like to get a good 3ish years out of my upgraded system before having to upgrade my entire graphics setup again. Considering that I wish to only use this for the oculus rift, the software and game demand side would be the only thing increasing every year.

So, the big question is for the oculus rift (1440p panel), should I buy another GTX 780 and SLI, or buy a 980 or even 980ti with the intention of SLIing in the future once the graphical demand catches up? All while considering differences such as Directx 12.

Thank you so much, I do understand this is a difficult question to answer but any help would be greatly appreciated.

Joey Martin

 

[DownTheSky]

980Ti is worth it. I'd suggest to wait a week for the AMD Fury release is and get the better one.

980 isn't worth it atm.

 

[exar333]

The 780 still sells for almost $300 on ebay for a benchmark. You would do better selling that and paying the difference for a 980Ti IMHO.

 

[Atomic Playboy]

I'd opt for the 980ti but not with the expectation of SLIing in the future. After all, you probably got the 780 with the intention of SLIing in the future and look where you are now. But if you are trying to future-proof your hardware choice, the 980ti is a better investment given the additional VRAM (not to mention the overall improvement in performance over a 980).

 

[Headfoot]

The 980 is a terrible buy. 980 Ti is way better for the money comparatively. Wait a week and see what AMD drops.

 

[RussianSensation]

The 780 still sells for almost $300 on ebay for a benchmark. You would do better selling that and paying the difference for a 980Ti IMHO.

Great advice. Sell 780 now and get 980Ti or Fiji cards. 780/780Ti/Titan are all getting crushed in the latest games while 980 is an overpriced card at $470.

OP, I suggest you look at this review:
http://www.guru3d.com/articles_pages/gigabyte_geforce_gtx_980_ti_g1_gaming_soc_review,13.html

780 SLI will lose to an after-market 980Ti, so if you are willing to buy a 2nd 780, don't even bother!

 

[Lepton87]

Unfortunately Kepler became crap. I don't know if it is because of the architecture or the drivers but it just doesn't run new games as well as it should. In older games my overclocked Titan was at least on par with 970. I had the 970 for a while and I tested it and compared the results to the titan and to Titan SLI. That was back when 980/970 was just released. Now I bought the Witcher 3 and both of my cards couldn't even provide playable frame-rates at 2560x1440. I just had a leak and one of my cards definitely burned I don't know about the other one. If it works I'm NOT going to buy another one because the newer the game the worse it runs it. I'm waiting for FURY because I'm very wary of buying the 980Ti and then the new GF arrives on 14nm and then 980Ti tanks like the Kepler going from excellent performance at 2560x1440 even with a single card to unplayable in SLI (The wither 3) The kepler architecture doesn't seem so deficient as to perform on par with Tahiti when it was 40% faster on release. It seems like planned obsolescence.