Hitman 2016 Performance Video Card Review

Hitman (2016) supports the new DirectX 12 API. We will take this game and find out if DX12 is faster than DX11 and what it may offer in this game and if it allows a better gameplay experience. We will also compare Himan performance between several video cards to find what is playable and how AMD vs. NV GPUs compare.

Introduction

Hitman (2016) was released on March 11th on the PC. This new Hitman game is the sixth entry in the Hitman series, but the prologue acts as a prequel to the series while the rest of the game takes place after the last Hitman game Hitman: Absolution. Hitman is an episodic action-adventure stealth game developed by IO Interactive and published by Square Enix.

Hitman has been heavily promoted by AMD, similar to Hitman: Absolution when it was introduced. We in fact used the last game quite a bit during its run in our gaming suite and performed a Performance and IQ Review on that previous game.

Article Image

Hitman Graphics and Options

The graphics definitely needs to be talked about in this game. Hitman (2016) runs on the Glacier 2 engine, which is the exact same engine that ran Hitman: Absolution in 2012. The engine has been modified slightly.

One modification the engine has received, which is very welcomed, and one of the main reasons we are testing this game today is that it supports DirectX 12 API at launch. Beyond that however, there doesn't seem to be anything noteworthy that is new about the graphics in this engine. In fact, some of the graphics options found in Hitman: Absolution are missing or not supported in Hitman.

Article Image Article Image

If you look at the graphics options found in Hitman: Absolution here and then the graphics options found in this new Hitman above you will find some missing options.

In Hitman (2016) you will find that MSAA is not there, you will find that Tessellation is not there, Global Illumination is not there, Bokeh Depth of Field is not there, and you will find only a basic form of Ambient Occlusion supported. Other than DX12 support the graphics seem to have taken a step back-wards from the previous game.

The only anti-aliasing options that exist are FXAA or SMAA or Off in this new game, plus a slider Super-Sampling option. The features that affect image quality are Level of Detail, Texture Quality, Texture Filter, SSAO (with either On or Off), Shadow Maps and Shadow Resolution. This is an overall very limited amount of visuals and control.

At the top of the settings menu you can select between DX11 or DX12 API. There is a memory override safeguard you can turn off if the resolution you want to run at isn't displayed. You see, this game will auto-detect your VRAM and only show the settings/resolution it deems is supportable by the amount of VRAM you have. You can thankfully disable this and run at anything you want. We did not encounter any limitations with the default "No" option.

We will be testing at the highest settings possible today, whether that be under DX11 or DX12. Note that the three main options that affected performance for us mostly in-game was Ambient Occlusion, Shadow Maps and Shadow Resolution. Shadow Maps does directly affect VRAM usage as well. However, Ambient Occlusion and Shadow Resolution can affect performance to a high degree in this game.

Patches

There have been a few patches for this game released since launch as well. The current patch is version 1.03. Most of the fixes have been load time or bug fixes, there doesn't or hasn't been anything related to image quality or game performance or DX12 stability or performance. We are using version 1.03 for our testing. The reason we just want to point out these facts is that because this game definitely needs a patch update for DX12, based on our findings stability and performance is lacking under DX12 in this game.

How We Tested

Since this game can run either DX11 or DX12 we are first going to find out which API is faster or better on each video card, then we evaluate what the highest playable settings are. In some cases this may mean DX12 API is better, and thus we find out what the playable settings are under DX12, in other cases DX11 may be better and thus we find the highest playable settings under DX11. Whichever is best for whichever video card is what we go with, so that means both APIs are tested and evaluated on every video card, even if we don't show what was playable for both.

For our apples-to-apples test, since we cannot record framerate over time for DX12 we are going to instead use the built-in benchmark and make average FPS bar graphs showing you the results. In our testing the built-in benchmark is not the best representation of performance in the game however. While it shows many scenes in the game, there are other scenarios we found to be more demanding in framerate that the built-in benchmark did not show. Therefore, the benchmark numbers are typically higher than what we actually recorded using in-game real-world run-throughs under DX11.

In other words, what might look playable by performance in the benchmark doesn't mean it actually will be in the actual game in some scenarios. If you set your graphics settings so that they are at 40 FPS on average in the benchmark, you will find some scenarios where the framerate actually drops to 30 FPS inside the game actually playing it. Therefore, don't set your playable performance by the benchmark, just use it as a suggestion or first guide, but you will find performance is slower in parts of the actual game than the benchmark depicts.

We do know that many readers do like solid framerate data, and we are using the benchmark in this instance to help us determine if DX12 is faster or slower than DX11 and because it is the only way to show framerate over time under DX12.