![]() ![]() DiRT Showdown speeds up from 70FPS average to a huge 115FPS, too. Shadow of Mordor sees a nice increase from 27FPS to 39FPS, while BioShock Infinite leaps from 46FPS to 82FPS. We may not have seen much SLI scaling at 1080p in Metro: Last Light, but at 1440p we jump from 35FPS average to 55FPS. Massive results, something that you simply can't complain about here with two GTX 960s in SLI. On the Ultra preset, the same performance jump is seen as we move from 59FPS to 113FPS average. On the single GTX 960 we were seeing a great average of 86FPS, but in SLI we jump right up to a huge 153FPS. The same great scaling in Battlefield 4 at 1920x1080 passes through to 2560x1440, where we start off with Battlefield 4 at Medium detail at 1440p. BioShock Infinite also jumped by a huge percentage, from 57FPS to a whopping 115FPS. Sniper Elite 3 went from 93FPS to 139FPS, while Tomb Raider scaled from 55FPS to 88FPS. Thief received a huge performance increase, jumping from 42FPS on the single GTX 960 to a huge 76FPS in SLI. GRID: Autosport and Metro: Last Light didn't see any additional performance. This is an incredible level of scaling, for one of the best looking games on the PC right now. Throwing another GTX 960 into the mix and enabling SLI saw this number jump by almost 100%, to a huge 127FPS average. Our single ZOTAC GeForce GTX 960 AMP! Edition performed well for the most part, where on the Ultra preset (without AA enabled) we were seeing an average of 70FPS. ![]() Starting with Battlefield 4, I didn't expect this level of performance. The 2GB of VRAM isn't enough, and swapping between the two cards is obviously hurting the outcome.Īny and all judgment on this SLI setup is going to come from the SLI scaling, overall performance, overall scaling at various resolutions, noise, price and power consumption. Futuremark does warn users that they require 3GB of VRAM to run the test at 4K, so this is why we're seeing a reduction in performance, rather than an increase. We see a total score of 1167 on the single GTX 960, while SLI actually hurts performance, dropping down to just 1070. Heaven, on the other hand, really enjoyed the second GTX 960, moving from 29FPS average to 54FPS average. The single GTX 960 scores 3468, while the SLI setup pushes out 5920. Unigine's Heaven didn't scale as well, but we still have an impressive jump from the 50FPS average to 86FPS average.Ĭhecking out the 1440p results, we can see that the scaling on 3DMark's Fire Strike Extreme benchmark wasn't as good as the 1080p run. I knew the synthetic benchmarks would receive a huge boost from the additional GPU, but just how well did they scale in SLI? 3DMark's Fire Strike at 1080p was a massive success, jumping from a total score of 6796 on the single GTX 960 to a huge 11,050 on the GTX 960 SLI setup. As for the detailed specifications, this is what we're running: We only recently built our new X99-powered system, something you can read about here. That's not to say that you won't purchase this card for it, but I would highly recommend that people looking to buy any of the GeForce GTX 960s to only get one if you have a 1080p capable monitor, and plan to not upgrade for a while. 2560x1440 (1440p) and 3840x2160 (4K) are going to be resolutions that this card won't be used for (mostly). Most people purchasing this card are only going to be gaming at up to 1080p, as the 128-bit memory bus and 2GB of VRAM is going to severely limit and hinder performance above that resolution. For now, let's get right into the synthetic benchmarks and see how two GeForce GTX 960s in SLI perform. This will take more time per review, as I'll have to invest time into actually physically playing the games, but it'll be worth it in the long run. Over time, I will be adding in new benchmarks and a new section that will concentrate solely on real-time gaming benchmarks. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |