Page 6 - Everest, HL2-Episode 1 & Crysis tests

Everest Home Edition
EVEREST Home Edition is a freeware hardware diagnostics and memory benchmarking solution for home PC users.

It offers accurate hardware information and diagnostics capabilities, including online features, memory benchmarks, hardware monitoring, and low-level hardware information. EVEREST Home Edition is optimized for Microsoft Windows XP and Windows Server 2003 operating systems, and it fully supports the XP look & feel.

Download

Everest is a tool that is growing slowly on me, so I decided to include the results. Synthetic numbers are simply the best way to show you the exact difference between timings and frequencies. Something that is very hard to show with games.

Charming and rather accurate reports are offered by this little benchmark suite especially for memory in this tool. Here we can see bandwidth grow once we alter all the goodness. As you can see with all three synthetic tests we achieve read bandwidth over 8100 MB/sec once overclocked which is very efficient at this FSB frequency. Write results are also fantastic.

Let's run a game and launch the same methodology used at it to see if we can perceive a performance difference.

Half-Life 2 - Episode One

We definitely wanted to include a game title as hey... we wanted to verify if faster memory actually has a real-world impact on the gaming performance.

Obviously is does. Now since we do not want the Graphics card to be a limiting factor in any way to objectively measure a memory timing or frequency difference we leave the CPU state in all times the same.

We disabled AA in the game also. Let's have a peek.

Episode 1 is configured at the best possible settings, everything is maxed out and enabled yet we left AA disabled as we did not want the graphics card to become a bottleneck.

We scale in resolutions from 1024x768 towards 1280x1024 and surprisingly enough, memory makes a measurable difference.

With a good timed 2GB 800MHz memory kit we achieve 150 frames per second in 10x7... yet we end up with 154 FPS on once we overclocked memory. Granted, these are very tiny difference, but every little bit and tweak in your system surely helps out if you look at it from a greater perspective.

Look closely at the Reaper memory, that's 3x more expensive memory folks.

Crysis Single Player CPU test

Graphically stunning, tactically challenging and always intensely immersive, Crysis sets player choice at the heart of its gameplay, with customizable tactical weaponry and adaptable armor allowing instant response to changing conditions. Judging from the downloadable demo, Crysis doesn't feel all that different from its predecessor, Far Cry. Both are set on an island. Both involve a latent (here in the demo, only briefly glimpsed) alien menace. Both bid you move more or less linearly through shaggy jungle areas, where the fact that you're progressing in a single direction is camouflaged by your ability to approach obstacles in your path any way you like. Think the "every time you play a situation yields radically different behaviors and results" approach in games like Rainbow Six Vegas or Gears of War except on more of a geographic scale.
 

Again we measure at 10x7 as any other resolution is a bottleneck for the system and do this with the hep of the Crysis CPU benchmark. We opt the CPU benchmark as there is a very strong correlation between system memory and the CPU. And significant change in memory will have an effect on the CPU and this framerate as well.

So again we see a tiny, yet measurable ~5% difference, it's roughly.