…and now I’m looking forward to catching up with the rest of you who’ve enjoyed the game over the last week while I was benchmarking. I have all of this hardware from which to choose. So, what am I going to put back together for my own Battlefield 3 gaming rig?
Processor-wise, it doesn’t really matter. I could pick an AMD FX, an Intel Core i3, or a Core i7 and get the same performance. I’ll probably throw a Core i5-2400 on a Z68-based motherboard just because the former rarely needs to be used for testing and I have spares of the latter. I’ll likely drop in a pair of GeForce GTX 570s for Ultra quality on a 24” (1920x1080) screen. And I’ll definitely use an SSD.
That’s not something I really tested for in this piece, but Dice does do a lot of object mesh and texture streaming during multiplayer levels. Minimizing level loads and the time you spend looking at a Saving prompt really encourages the use of faster storage.
One last thing that needs to be mentioned: I made it about five levels into the single-player campaign before buckling down to benchmark. I have a lot of data here based on that 90-second snapshot where there is no running, jumping, or magazine-emptying combat.
A distinction even has to be made between this single-player experience and what you’re almost certain to see in a multi-player map. I’ve seen benchmarks from the multiplayer beta where there’s more of a hit incurred dropping down a given processor vendor’s stack. What’s most clear there and here, however, is processor performance takes a back seat to graphics potential.
But this sequence is friggin consistent. So, when you compare frame rates, keep relative performance in mind, and not the absolute numbers attained here, which are guaranteed to range both up and down, depending on what you’re doing in Battlefield 3.
AMD And Nvidia: Two Disappointments
I had my own issues in testing for this story, and they are summed up in two points: AMD’s CrossFire support and Nvidia’s image quality on DirectX 10-based cards.
AMD first. Apparently, prior to launch, the company had everything running swimmingly. We do know that Dice pushed out a 222 MB v.1.01 patch prior to Battlefield 3 going live, though, and perhaps it broke something. Regardless, shortly before publication, AMD confirmed it had replicated my problems with CrossFire and was trying to figure out what happened. Should the company fix this issue with haste, maintaining excellent scaling, it’ll emerge a favorite for its exceptional performance.
I’m not sure what’s going on with Nvidia’s Low Terrain Quality setting in DirectX 10, but it looks really bad. AMD’s Radeon HD 4870 didn’t exhibit the same artifacts, so I have to assume this is fixable in software. Until that happens, I’d be inclined to stay away from anything in the GeForce 200 series or older. Fortunately, the GeForce GTX 400- and 500-series cards look great, run well, and take a much smaller hit when you apply MSAA than AMD’s boards.
A special thanks to the Tom's Hardware readers who engaged with me on Twitter to help shape the direction of this story as I tested. If want to play a more active role in the projects we tackle, make suggestions on the tests we run, or simply share your enthusiast for hardware technology, please join me!
- Battlefield 3 Reinvigorates PC Gaming
- Test Setup And Sequence
- Benchmark Results: Nvidia Graphics Cards, High Quality
- Benchmark Results: Nvidia Graphics Cards, Low Quality
- Benchmark Results: Nvidia Graphics Cards, Post-Process And MSAA
- Benchmark Results: Nvidia Graphics Cards, SLI
- Benchmark Results: Nvidia Graphics Cards, What Do I Need For Ultra Quality?
- Benchmark Results: AMD Graphics Cards, High Quality
- Benchmark Results: AMD Graphics Cards, Low Quality
- Benchmark Results: AMD Graphics Cards, Post-Process And MSAA
- Benchmark Results: AMD Graphics Cards, CrossFire
- Benchmark Results: AMD Graphics Cards, What Do I Need For Ultra Quality?
- Benchmark Results: CPU Scaling
- Battlefield 3 Is Good For PC Gaming….