Last week we introduced you to the Evetech Lagzatron 3000, the beauty that would serve as our new test bench for testing PC hardware. Before starting with any of that we needed a standard, vanilla benchmark to have something to compare future tests to, and prove that we kind of know what the hell we’re doing. So how did everything turn out?
Quite well actually. We chose a few games that usually do a good job of pushing PC hardware to its limits, cranked up the resolution and hoped nothing caught alight. Nothing did, and the Lagzatron impressed in more than a few ways, especially when trying to push out 4K resolutions.
Firstly, some ground rules…
Before anything, we had to establish some set rules that will be held up in the future. We’re not testing graphics cards exclusively from Nvidia, which means we had to throw out any Nvidia-specific features. This includes PhysX, TXAA and HBAO+. The same goes for AMD exclusive features such as Mantle, but we didn’t really have to worry here considering AMD loves sharing their software with everyone.
In every test we conducted for the benchmark, and for every test going forward, these settings will remain off. This just keeps the playing field even in the fairest way possible. It comes in handy when comparing Nvidia and AMD cards directly, which is definitely a big talking point nowadays.
We also chose three different resolutions to run our games at, with the visual settings not changing between tests. So if it’s running at Ultra at 1080p, it’s being forced to run at Ultra at 4K as well. Again, just keeping things constant for reliable test scores. Each test is also run at least three times to ensure consistency, because we all know that odd things happen now and then. With that out of the way, we’re ready to see how well the Lagzatron did.
3DMark Fire Strike
Futuremark are the go to guys for PC benchmarks. Their latest test software, 3DMark, features numerous different tests for all sorts of devices, including smartphones. For a PC such as the Lagzatron, we fired up the Fire Strike benchmark test to put the CPU and GPU through their paces. The score was pretty impressive, but more importantly it gives us something to compare future tests to.
Crysis 3 is definitely not a benchmark in terms of gameplay, but there are few other titles that push a PC to its limits like this. The game is technically impressive, and cranking it up to higher resolutions just made the smaller details shine brighter. It also brought the Lagzatron to its knees, which isn’t all that surprising. For testing we ran through the first three minutes of “Welcome to the Jungle”, running with FXAA anti-aliasing and everything else maxed out.
Metro: Last Light
Metro: 2033 was notorious for being one of the most unoptimized PC titles ever released, making it perfect for testing out new hardware on the market. Last Light isn’t nearly as bad, but it’s still an extremely demanding game in its own right. The game comes with a very nice benchmarking tool that was used for the tests, with all settings at their highest value and PhysX off.
Ubisoft’s new IP earned itself a bad rep with PC gamers earlier this year, after it ran pretty badly with nearly every hardware configuration out there. Numerous patches have alleviated some of the issues, but the game is still a VRAM hungry animal. That makes it a perfect test for a lot of graphics cards, especially new ones that will feature insane amounts of memory. Texture settings were kept to high here, with FXAA anti-aliasing and all other settings set to Ultra. Of course, no HBAO+ either.
Batman: Arkham Origins
Batman: Arkham Origins is nearly a year old, and it’s really Arkham Knight that’s going to be asking big questions of your PC. Regardless, the open-world title has a neat little benchmark tool that is great for testing out hardware, even if the engine seems to be fairly forgiving. That said, pushing out 4K resolutions wasn’t exactly a walk in the park. Again, we used FXAA anti-aliasing and as many DX 11 enhanced features we could find.
These are the baseline tests we’re going to use to measure hardware performance going forward, and we could add to them in the future. As hardware advances so do game engines, which means our benchmark test could change over time. Every time we do make a change, we’ll update the vanilla scores you see here as well.
With that said, I think our vanilla setup did rather well with what we threw at it. I’m excited to start testing some other hardware in the Lagzatron 3000, and you should expect some reviews very, very soon.
Once again, thanks to Evetech.co.za for supplying us with the hardware necessary to do this, and Samsung for the beautiful UHD monitor.
Last Updated: August 4, 2014