Moving away from CrystalMark 2004R2 for benchmarks

2gb_ram_q1p_umpcPosting has been lighter than usual this morning and I wanted to provide the reason. I’ve spent the last few hours running CrystalMark 2004R2 benchmarks on the Q1P since the 2GB RAM upgrade was successful. After seeing some results, I’ve decided this isn’t the benchmark tool I want to use for this test; let me explain at a high level.

I’m trying to be very meticulous in my testing and in the end, I expect you’ll see six different result sets from my Samsung Q1P due to several specific configurations:

  • 2GB of RAM with two registry tweaks and 2GB of RAM without the tweaks
  • Same two scenarios with 1 GB of RAM
  • Same two scenarios with 512 MB of RAM (just for comparison)

I’ve run most of these scenarios and unfortunately, the results are not different with any statistical significance which tells me I need to use a different benchmarking application. Or it could mean that Vista runs equally well on the three different memory configurations. ;) I believe that CrystalMark is a good program, but it’s not the right tool for this particular task. More to follow….


Comments have been disabled for this post