• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

GTX 780ti in 2016

Permabanned
Joined
12 Sep 2013
Posts
9,221
Location
Knowhere
A few days ago Gamers Nexus uploaded a video where they retested the gaming performance of the Kepler flagship to compare with the current cards, I couldn't find a 780ti owners thread to put it in so here it is.

 
Is losing to RX470, what are you talking about? Let alone it doesn't support DX12 or Vulcan

It's certainly not ageing well, As he say's in the video it does okay in older titles but it suffers in the new gen console titles. It'll be interesting to see if the same holds true for Maxwell but I don't think it will be as bad for them as they're an architecture that was developed alongside the new gen.
 
Nope, the closest the 780ti can get to the 1060 is in Metro and even then it lost by a small margin. Watch the video and try and argue otherwise.

Thing is he is running it at 928MHz and from the way he talks sounds like he is limiting GPU Boost behaviour to the reference boost clock - aside from some barrel scraping jobbies most of the 780tis people would own will have around 1085Mhz boost clock and running around 1100MHz actual out the box before any overclocking. Once again Kepler is under-represented in benchmarks compared to what actual users will be seeing (this also includes how actual users will see their cards versus what actual users of other cards will be seeing). Though that increase in clockspeeds alone isn't going to dramatically change the story in terms of those benchmarks.

I also find their Doom results a bit strange - 64fps v 85fps - in my own testing 1080p ultra with my 780 clocked to a little faster than a reference 780ti it was only slightly lagging behind a RX480 by approx 1.5% even with the clock backed off a bit to what would be 780ti performance it wouldn't make the cards around 30% different in speeds - that said I've found Kepler performance in Doom very strange with a lot of inconsistency between setups and resolutions so its quite possible they are seeing very different performance to me.
 
Last edited:
running mine at 1100 on the core no idea what boost is and it seems to run all games great at 1080p

I don't get the fascination reviewers/benchmarkers have for running GPU Boost 2.0 cards at very low boost clocks - often if you can replicate the results you only get the numbers they do if you artificially limit the boost clock to the exact on paper clock - most of the other cards are benchmarked close or exactly what the end user would see - ok some 290 cards are a little down on realworld clocks but its only like a 2.5% difference in performance while the Kepler cards are almost always benchmarked 10, 15 or even 20% (extreme case for a B1 card) below what most owners will see out the box.

I get doing it for the older GPU Boost 1.0 cards as often they categorically would not hold their boost clocks and drop to the on paper clock after 30-60 minutes but GPU Boost 2.0 cards extremely rarely if ever exhibit that behaviour.
 
My ancient GTX 690s can get well over 1200mhz with a bit of voltage and hold it and that is on air too.:D

That isn't with the out the box settings though :P

I'm being a bit pedantic here as years of running 700 series cards has shown me a very different out the box picture compared to the mainstream benchmark site results.
 
Not really surprising - nVidia cards have always been about getting the best performance for the here and now (even when they failed to achieve it - FX series for instance). Back in the day you were lucky to get more than about 18 months out of a GPU before it was rendered obsolete at the top end.
 
running mine at 1100 on the core no idea what boost is and it seems to run all games great at 1080p

Same here. Mine runs at 1121 and boosts to 1251, dropping to 1238 if it exceeds 65C when the card runs at 100%.

No complaints at 1080P so far.
 
I must have a super computer as with bf1 i get around 70fps average with a 780 @1920x1200.

Disclaimer, I have not played yet with the newest patch or drivers.
 
I have a Gigabyte 780ti, and no real complaints at 1080p.

Still getting the upgrade itch however ;)

I am getting mixed messages though about the upgrade path. would something like a 480 be a worthwhile upgrade? I used to think not but not sure now - or do I need to stretch into 1070 territory to see a decent boost?
 
Back
Top Bottom