Date: Wednesday, July 25, 2018
The premise is simple, we are going to see how GPUs perform from generation to generation over the course of the last five years in games. This is a new and unique review format we have never done before. Quite simply there is so much information that putting it all into one big review would just be too messy and hard to understand. Instead we are going to narrow the focus by separating this review into multiple parts. At the end of this series of articles we will pull it all together and cross-compare GPUs and gaming performance.
Let’s say you are holding onto that older video card, waiting for the right time to make the big upgrade. However, you don’t really know what kind of a performance gain you are really going to get in games. Or let’s say you just want to take a look back and justify all those upgrades you’ve made in the past and track the performance gains over the years. Or, let’s say you want to see if performance gains in GPUs are getting worse or better over time.
We are on a journey to find out the real-world performance gains from GPU "generation to generation" over the course of many years. We want to find out how much of a gain there has been, and has this performance advantage been shrinking or growing over time from generation to generation, i.e. have the gains been less as time goes on or have the advantages been greater?
Certainly when the next-gen cards come out, this will give us a solid and fairly up to date data points instead of relying on data that is years old.
To do this we are going to compare NVIDIA GPUs and AMD GPUs over the span of five years. Five years is a good round number to look back on and takes us to the year 2013. To go any earlier would mean unplayable performance in today’s games. Though we know gamers can and do hold on to older video cards waiting for the big upgrade, most of our hardcore hardware enthusiast gamers who read our content most likely upgrade every 2-3 years, at the least, if not every year!
Part 1 today is focused solely on NVIDIA GPUs from 2013 to 2018. Part 2 will focus on adding the GeForce GTX 980 Ti and GeForce GTX 1080 Ti to the mix. Then Part 3 will focus on AMD GPUs "generation" to generation" from 2013 to 2018.
We put "generation" in quotes above because we realize and understand that technically some of these GPUs are the same architecture, consisting of re-fresh products or evolved products based on the same architecture. We simply cannot include every single video card under the sun from NVIDIA between 2013-2018, that would next to impossible for us since we are using actual gaming instead of quickie canned benchmarks.
To make this even more interesting we are going to use games that span the same years, 2013-2018. We have chosen at least two games per year for each video card launch date. Please understand there are a lot of games out there, and that your game you wanted to see may not be explored in this review.
We had to be choosy about the games we picked. We decided to pick games that were either the most popular during that year, or highly anticipated, played and well received, or graphically intense. We also wanted to choose games based on either being released before a card was lunched, or after a card was launched so both are represented for each year and card release. We wanted these to line-up with the card launches so that way each video card got its best chance to perform in the game that was released the same year as the video card.
The games we have chosen are, from earliest to latest: Crysis 3 2013, Tomb Raider 2013, Grand Theft Auto V 2013, Far Cry 4 2014, The Witcher 3 2015, Fallout 4 2015, Rise of the Tomb Raider 2016, DOOM 2016, Deus EX Mankind Divided 2016, Battlefield 1 2016, Sniper Elite 4 2017, Mass Effect Andromeda 2017, Kingdom Come Deliverance 2018, and Far Cry 5 2018.
Every video card mentioned above is being tested in every game above at 1440p. As you can see, that makes three video cards times fourteen games, with two run-throughs per-game which equals 84 run-throughs of data! Just for these first three NVIDIA GPUs.
We are testing both high levels of graphics settings and low levels of graphics settings in each game. We wanted to include a low game setting comparison because of the GeForce GTX 780. We needed to give it a chance in the newer games so that it does not bottleneck and we can get an accurate performance comparison. Otherwise, the GPU bottlenecks in modern games quite easily.
On the other hand, we also must test a high graphics setting to give the GeForce GTX 1080 some kind of a challenge in the older games. Therefore, both are tested on each game page.
An important consideration for this review is the fact that both the games and video cards have all been out for years now. This means all the games being used are patched up to their latest patches and have received all the patches and optimizations they are ever going to receive in their lifetime for the most part. Likewise, we are using the latest video card drivers, which also have now been optimized as much as they are ever going to be optimized for each video card.
That means now in 2018 we are experiencing the absolute best each video card has to offer in terms of performance in games, both the games and drivers are as optimized as they are ever going to get. This is the pinnacle of performance.
We want to give a brief run-down of the specifications of the video cards being evaluated to remind you what architectures are being tested.
GeForce GTX 780 - The GeForce GTX 780 was launched on May 23rd, 2013 with an MSRP of $649. The GeForce GTX 780 is based on the Kepler architecture on a 28nm process. This is actually an evolution of the Kepler architecture found previously in the GeForce GTX 680. Therefore, this is a refreshed Kepler architecture. The GeForce GTX 780 consists of 2304 CUDA Cores, 192 Texture Units and 48 ROPs. It runs at a baseclock of 863MHz and a boost clock of 900MHz. It contains 3GB of GDDR5 on a 384-bit memory bus running at 6GHz. This gives It 288GB/sec of memory bandwidth. The TDP is 250W. The obvious downside to this video card in today’s games is the small 3GB of VRAM.
GeForce GTX 980 - The GeForce GTX 980 was launched on September 18th, 2014 with an MSRP of $549. The GeForce GTX 980 is based on the Maxwell architecture on a 28nm process. The GeForce GTX 980 consists of 2048 CUDA Cores, 128 Texture Units and 64 ROPs. It runs at a baseclock of 1126MHz and a boost clock of 1216MHz. It contains 4GB of GDDR5 on a 256-bit memory bus running at 7GHz. This gives It 224GB/sec of memory bandwidth. The TDP is 165W.
GeForce GTX 1080 - The GeForce GTX 1080 was launched on May 17th, 2016 with an MSRP of $599 or $699 for the Founders Edition. The GeForce GTX 1080 is based on the Pascal architecture on a 16nm FinFET process. The Pascal architecture is an evolution or refresh of the Maxwell architecture, with some tweaks and major clock uplifts. The GeForce GTX 1080 consists of 2560 CUDA Cores, 160 Texture Units and 64 ROPs. It runs at a baseclock of 1607MHz and a boost clock of 1733MHz. It contains 8GB of GDDR5X on a 256-bit memory bus running at 10GHz. This gives It 320GB/sec of memory bandwidth. The TDP is 180W. To this date, this is the still the latest generation of GPU and next to the fastest gaming GPU from NVIDIA.