Performance Games - Dota 2 / Grand Theft Auto V / Ghost Recon: Wildlands

Performance - DOTA 2 (DX11)

We'll start with a massive online multiplayer game at 1920x1080, our choice is DOTA 2. While the majority of the time players might be running around alone, slaying creeps and heroes alike, the worst-case scenario in a game of Dota 2 usually is a large-scale teamfight. This fight pushes the CPU and GPU to the maximum: Massive amounts of hero models, hats, particles and spells flying around and dropping the framerate.

We also apply:

cl_showfps 2
fps_max 0 //Dota 2 defaults to a 120 fps cap, this gets rid of that
dota_spectator_mode 0
dota_spectator_hero_index 7
host_timescale 0.15
timedemo_start 58200
timedemo_end 59000
And run a timedemo.

We use a sufficient demo file proved to match 3061101068, the first game of the BO5 Grand Finals of Elimination Mode 3.0 between Evil Geniuses and DC (formerly Onyx). There are heavier workloads available but this is a very intense scene and since EM3 was a Moonduck tournament there’s zero harm using it. We render at DX11.

For DOTA 2, just one resolution. The game, as we already discussed in the past, is extremely susceptible towards CPU and memory changes when the game is in a CPU limited environment. The quick to render game was tested pretty much at medium (integrated graphics) quality settings. Hence the difference is visible, bigtime.

Grand Theft Auto V (DX11)

For CPU and motherboard reviews we have added IGP performance. This is a new set of tests, on this page we'll look at Grand Theft Auto V. The game has been set at 'normal' settings and thus is rendering over the integrated graphics unit inside the processor/APU.

 
GTA-V is well-behaved and runs on many platforms, performance, however, is difficult to emulate. If you click the image above, you can mimic the settings and try for yourself. We use the internal benchmark mode and note down scene run 4 results.

GTA-V is a bit more GPU bound, ergo you'll see difference decrease quickly We applied rather CPU bound quality settings by keeping the image quality settings are kept at 'normal'. As you can see, there is still a measurable difference, let's call it 5%. Would we increase the graphics quality settings to good quality, the CPU and memory would be less of a bottleneck and the GPU would become the bottleneck again (which is the case in 9 out of 10 systems). Then the perf fifferential would be much smaller.

Performance - Games - GeForce GTX 1080 - Ghost Recon: Wildlands (2017)

Tom Clancy's Ghost Recon Wildlands is an open world tactical shooter video game developed by Ubisoft Paris. It is the tenth installment in the Tom Clancy's Ghost Recon franchise and is the first Ghost Recon game to feature an open world environment. The game moves away from the futuristic setting introduced in Tom Clancy's Ghost Recon Advanced Warfighter and instead feature a setting similar to the original Tom Clancy's Ghost Recon. Ubisoft described it as one of the biggest open world games that they have ever published, with the game world including a wide variety of environments such as mountains, forests, deserts and salt flats.

For Ghost Recon: Wildlands I opted for proper quality settings, very high quality, the way we test and stress graphics cards in our reviews. This is pretty much the most realistic performance difference measured that you'll notice between 2133 MHz memory and fast 3600 MHz clocked memory. At 1920x1080 you'd be looking at a whopping 1 FPS differential. That, my friends, is the reality with your PC, it is the sum of all parts that define your game performance, the system component with the biggest bottleneck will always primarily define your framerate and, as stated, in 9 out of 10 cases this always will be your GPU.