May the FPS be with you! We test Star Wars: Battlefront's graphics performance on different hardware. Our results provide some good news: older graphics cards do just fine fighting the Galactic Empire.
This isn't a game review. Star Wars: Battlefront doesn't have a single-player campaign anyway, and debating the pros and cons of the title's map design is a futile effort. There are a few single-player and co-op missions that ease new players into the game. But there's certainly not enough content to keep an offline player engaged for long.
We'll leave the opinions about game play to the software reviewers and concentrate our efforts on the playability of Star Wars using different hardware configurations. When it comes to graphics, EA scores major points. Its proprietary Frostbite 3 engine now employs physics-based rendering (PBR), which is used to great effect.
Using physics to model the interactions between lighting, surface structures and materials is popular these days. This approach might trivialize the underlying laws, but it's undeniable that the end result is a more realistic visual experience.
This way, the developers kill two birds with one stone, simplifying programming/minimizing potential errors and dialing back hardware requirements, since optimized libraries take care of the heavy lifting.
A nice side effect is the lighting model (Global Illumination) that everything's based on. It looks great on even the lowest-end graphics cards.
The vegetation and character models from the Star Wars universe are very detailed, owing to great textures and ample geometry.
As opposed to traditional 3D scans, photogrammetry uses pictures taken from different angles. A specialized software package computes the necessary meshes and vertices, representing the corner points of a complex object's individual surfaces. That's critical to the finished product's visual quality and performance.
Additional algorithms can be used to automatically make an object simpler. For instance, unnecessary or invalid surfaces can be deleted, and vertices that are very close to each other can be merged. As a (welcome) side effect of this approach, textures can be mapped to vertices exactly since they're generated at the same time. These highly-optimized models yield great performance, even if the environments are fairly complex.
Like most current titles, Star Wars Battlefront doesn't support Mantle or DirectX 12. However, the game's graphics engine is tuned so well that it still runs smoothly and looks good on older hardware.
MORE: Best Graphics Cards For The Money
MORE: All Graphics Content
Test Scene Selection
We're using Endor for our tests. It's the environment that poses the biggest challenge to your graphics subsystem. Running around (on foot) in the jungle vegetation's tight spaces mimics what you'll be doing through much of the game.
The Survival mission, specifically, allows us to reproduce the same sequence over and over. Doing so requires the use of a cheat that makes us completely invisible to the AI. Movement is executed exclusively through the keyboard with the help of a pre-recorded macro. This makes every run almost identical, though. Each benchmark pass takes 45 seconds. After that, we hit the limitations of our macro-based method. Each number you see is the product of three runs averaged together, giving us reliable measurements.
For most of the game, your frame rates should be significantly higher than our test results. The idea is that planning for a worst-case scenario by seriously taxing our hardware guarantees you'll get a more enjoyable experience when it comes time to play Star Wars for real.
Let's take a quick look at the test system before we get to the results.
|Test System and Settings|
|Test System||Intel Core i7-5930K at 4.2GHz + Alphacool water cooler|
Crucial Ballistix Sport, 4 x 4GB DDR4-2400
MSI X99S XPower AC
Crucial MX200 500GB SSD (system), Corsair LS 960 960GB (applications + data, storage)
be quiet! Dark Power Pro 850W
AMD FX-8350, be quiet! Dark Rock Pro 3
8GB Kingston HyperX DDR3-1600
MSI 970 Gaming
Corsair LS 960 960GB
be quiet! Dark Power Pro 550W
Windows 10 Pro (All Updates)
|Drivers||AMD: Catalyst 15.11.1 Beta|
Nvidia: ForceWare 359.00 (Game Ready)
Intel: 18.104.22.16800 (15.40.10)
|1920x1080, Ultra Preset|
3840x2160, Ultra Preset
Note: We're not reporting fractions of our average frame-per-second results, since it's just not possible to reproduce them to that degree of accuracy. Consequently, were using the average of three test runs, rounded down the next-lowest FPS measurement.
Results: 45 Graphics Cards And iGPUs At Full HD
Ultra Preset For (Almost) Everyone
Since our Endor test scene is one of the most challenging you'll encounter, we'd expect average frame rates to be much higher in other maps. On the next page, we’re also going to show how the different presets affect performance and where (good) playability ends.
We were very disappointed that SLI and CrossFire weren’t supported at launch or at the time of this test. After digging into our bag of tricks, we came up with a way to have the game use Battlefield 4’s profile, resulting in a touch of alternate frame rendering (AFR). Unfortunately, even this ended up not really being usable.
AMD's Radeon graphics cards perform very well across every generation, demonstrating the value of optimizing drivers before a game launches and even taking part in some of the development process. We’ll see a bit later how Fiji catches up with its competition when the going gets roughest.
Generally speaking, everyone who wants to play this game should be able to run it smoothly, including owners of entry-level GPUs and APUs. If you own a more powerful graphics card, we strongly recommend a quad-core host processor for the online maps. Running it at high clock rates doesn't hurt either. Star Wars: Battlefront might not scale linearly as you add on-die resources, but six or eight cores still yield smaller performance gains.
All of this makes Star Wars: Battlefront a good example of a well-optimized title that successfully compromises between eye candy and hardware scaling. It also does a great job running on lower-end configurations.
Results: UHD And Quality Presets
UHD With Ultra Preset
Since we’re testing the worst-case scenario here, every graphics card down to AMD's Radeon R9 390X and Nvidia's GeForce GTX 980 delivers acceptable frame rates.
The quality preset scaling results below might be worth a look, though, if you're looking for a smoother gaming experience. On the lowest graphics preset, even a Tahiti-based AMD Radeon R9 280X works at 4K. A stock Nvidia GeForce GTX 770 with its 2GB of GDDR5 can't always keep up, on the other hand.
Scaling With Different Graphics Presets
How much performance can you add by lowering the graphics quality settings? Regardless of resolution, performance roughly doubles when you go from the Ultra quality preset to Low.
The subjective differences between presets is smaller than you might think. Moreover, they're certainly smaller than you'd expect given the impressive performance bars. Endor's vegetation, including its flowing water, does look a bit flatter and more static, but it's certainly not like the graphics quality takes much of a hit.
Fortunately, the differences are just as small on Tatooine. What jumps out are the simpler shadows and reduced depth. The rest ceases to matter once your online enemies enter the game.
Whether you end up liking Star Wars Battlefront will depend on opinion of its gameplay, including the fact that it’s almost exclusively a multi-player affair. If you end up setting Star Wars aside, we can guarantee it won't be because of graphics quality or performance. This isn’t the place or time to lament missed opportunities for a stellar campaign. We're here evaluating the title's technical aspects, which EA did a great job with.
MORE: Best Graphics Cards For The Money
MORE: All Graphics Content
Igor Wallossek is a Senior Contributing Editor for Tom's Hardware Germany, covering CPUs and Graphics.