Batman: Arkham Asylum: GPUs, CPUs, And PhysX Performance

CPU Benchmarks: PhysX Enabled

We were somewhat puzzled by the results we saw when we benchmarked the Radeon cards at the different PhysX settings. Strangely enough, the minimum and average frame rates stayed static between the Normal and High PhysX settings, even though the GeForce cards showed a large performance hit for the minimum frame rate.

We therefore ran a couple of tests using Radeon cards with PhysX enabled at different CPU speeds:

This would seem to make sense, because the frame rates go down with lower clock rates. Here's the weird part, though. We took note of CPU utilization during a specific part of the game benchmark, a PhysX-heavy segment where a guard is walking through smoke. Results were repeatable and consistent over several runs:

Why is CPU utilization lower when PhysX is enabled? And why is CPU utilization so low at all? If the CPU is bottlenecking the PhysX calculations, shouldn't the increased load be pushing the CPU to its limits?

Maybe we can make more sense of this if we look at how the CPU is handling threads:

Rather than clearing things up, the results of this testing have only left us more puzzled. We can recall that the PhysX API is supposed to be optimized for multiple threads, yet on the Core i7, only a single thread seems to be stressed when PhysX is cranked up. We're trying to get clarification from the developers at Rocksteady about this phenomenon--it's almost as though the game is artificially capping performance at a set level, and is then using only the CPU resources it needs to reach that level. On the Core i7, PhysX is using fewer resources than it does on the Phenom II. This would make sense if there was an artificial performance cap, as the i7 has shown to outperform AMD's architecture. However, the benchmarks show us a correlation between actual performance and raw CPU speed on either CPU.