NVIDIA GeForce GTX 760 Review: The New Enthusiast Kepler
by Ryan Smith on June 25, 2013 9:00 AM ESTFar Cry 3
The next game in our benchmark suite is Far Cry 3, Ubisoft’s island-jungle action game. A lot like our other jungle game Crysis, Far Cry 3 can be quite tough on GPUs, especially with MSAA and improved alpha-to-coverage checking thrown into the mix. On the other hand it’s still a bit of a pig on the CPU side, and seemingly inexplicably we’ve found that it doesn’t play well with HyperThreading on our testbed, making this the only game we’ve ever had to disable HT for to maximize our framerates.
With Far Cry 3 we’re back to a game where the GTX 760 has a clear lead, surpassing the 7950B by an unexpected 28% at our highest 1080p settings. However 41.5fps is less than idea for an action game like Far Cry 3, so we have to drop the MSAA to take a crack at 60fps. Even then, the GTX 760 leads by 14%.
This also happens to be another great example of the ROP difference between GTX 760 and GTX 660 Ti. With MSAA enabled the GTX 760 takes a clear lead, but without MSAA the two cards are practically tied.
Meanwhile Far Cry 3 is also another strong example of the performance gap between the GTX 760 and cards 2-3 years old. GTX 760 leaves those older cards in the dust by anywhere between 77% for the GTX 560 Ti and 6870, up to 138% for the GTX 460 1GB. CPU performance may have plateaued over the last half-decade, but thanks to the embarrassingly parallel nature of graphics, we continue to see massive gains over a whole generation.
Facebook
Twitter
RSS
110 Comments
View All Comments
Ryan Smith - Tuesday, June 25, 2013 - link
Going by my chart, I have the GTX 760 winning in 7 of 10 games (all but DiRT, Hitman, and Crysis 1) at our highest 1080p quality settings, which is where I'm focusing on for a card this expensive. Of those magnitude matters; most of those GTX 760 wins are in the double digits, so the average does indeed end up being 8%As for frametimes, the idea is that we would normally include them. That said this review left us crunched for time; I would have likely needed to drop the Fermi cards to make time. With that in mind, there's absolutely nothing interesting going on with single-GPU frametimes right now with the games we use. The only place NVIDIA still differ are under multi-GPU scenarios.
Zstream - Tuesday, June 25, 2013 - link
Shouldn't we be using the median instead of average?ewood - Tuesday, June 25, 2013 - link
median, mean and mode all show very different things. you should have all three to draw detailed conclusions, however if only one is available i would personally prefer it be the mean.ShieTar - Thursday, June 27, 2013 - link
With a sample size of 10, the median would not be a very helpful information. To be honest, the mean is not all the important either. The distribution of performances is not all that random when comparing nVidia and AMD, but rather there are significant preferences for one architecture by each game.So everybody on the lookout for a new card should mainly be checking for a benchmark specifically on the game he/she spends the most time with. For this reason, I would love to see the benchmarks on Anandtech include the name of the engine for each game (if it is a licensed one), and maybe provide some handy reference to figure out what other games use the same reference.
And personally, as a player who does not play reaction-based games like shooters or racers a lot, I would love AT to re-introduce a BioWare and/or Blizzard title back into their benchmark-zoo. Even if those are not extremely new or demanding, I think they still have a high importance for a large number of players who don't care much to shoot virtual people in their virtual faces.
MarcVenice - Tuesday, June 25, 2013 - link
I see, I counted Sleeping Dogs as a win for the HD 7950 as well, considering the minimum fps is a bit higher. Thanks for the reply, I agree that if all is well with frametimes in a certain game, fraps is still a good way to measure raw rendering power.JeBarr - Tuesday, June 25, 2013 - link
Most multi-GPU users are reporting that the frame time issue mostly exists for 2-way SLI and Quad SLI. It seems that 3-way SLI or Dual GPU single slot SLI is the way to go for gamers concerned about the stutters. I'm not sure about 4-way SLI though, since I don't bother with it anymore. I can however, confirm that in my personal experience a single GPU or 3-way SLI is mostly unaffected.draknon - Tuesday, June 25, 2013 - link
This seems like a good spot to upgrade from my 460gtxEzioAs - Tuesday, June 25, 2013 - link
Yeah, it is. My original plan was to get the GTX 760, but Nvidia delayed it and I wasn't going to wait anymore, so I went and bought the GTX 660.DanNeely - Tuesday, June 25, 2013 - link
I'm tempted to upgrade to it from my 560 when I built a Haswell box later this summer and put the difference between it and a higher end card toward a better SSD, etc and then get a top end Maxwell based card next year.I'm a bit concerned about ending up in the same trap I did last time though. I bought the 560 as a stopgap replacement in Jan 2012 after stupidity killed my 5870, with the intent of upgrading to a GK100 based card in half a year or so only to have nVidia fumble its top end launch.
omarccx - Tuesday, June 25, 2013 - link
It seems like an even greater upgrade from my HD4000. :x