Along with other hardware websites around the internet, we at Rage3D.com have been given the opportunity to take an early look into the very soon to be released AquaMark3 benchmark application.
It is clear that Massive is aiming AquaMark3 directly at FutureMark and it's 3DMark03 synthetic benchmark app; Massive Development tagged this release as the "Reality Benchmark", an obvious shot. Unlike Futuremark however, Massive has taken the approach of building their benchmark app by using an actual, production quality game engine called the "krass™ Engine". Granted, the krass engine isn't exactly getting the same level of attention from game developers as the Quake3 Engine or the Unreal Engine gets, but it does at least represent gaming in a fairly realistic situation.
Having said that, many people (including myself) who have been dumping on 3DMark03 lately must admit that even though 3DMark03 doesn't represent any specific game, you don't have far to go to find the benchmarks and conclusions that confirm that what 3DMark03 has been predicting for seven months is indeed accurate in todays real DX9 games. So where does AquaMark3 fit? Can an engine used in current
games also be used as a point of reference for future games and future hardware? It's hard to say. Massive has indeed created a powerful engine with the krass engine. It sports many DX9 features and throws around an awful lot of triangles and high-res textures, not to mention the fact that they've also upped the vertex and polygon count specifically to try and predict future game performance. Games, even games based on the same game engine and same API do many things differently though. One system may scream through one game, only to just get by with acceptable performance on another. Ideally AquaMark3 would be used as a single point of reference in a review which has many such points. Whether AquaMark3 makes that point valid and proves itself worthy of our time end effort is exactly what we are going to try and determine here in this review.
Since this is a review of a benchmark application, I've decided to make use of as many cards and as many platforms as I have access to. This should give us a general idea as to what kind of information AquaMark3 gives and if it can be considered a reliable and useful tool for the evaluation of system performance.
A quick caveat first: as much as I would like to provide it, those looking solely for an ATI vs nVidia numbers fight will be disappointed here. We are
Rage3D and we don't deny that we are
an ATI support/fansite. That fact doesn't exactly put us at the top of nVidia's review-sample list and, since we don't have the resources available to purchase a 5900 Ultra of our own just yet, I can only show you what various ATI graphics cards will perform like across different systems in AquaMark3.