Is there a reason why you are setting NvAGP=2? Do your benchmark results change if you remove NvAGP from your X configuration?
Is this the VTK that you are referring to?
If so please provide the following:
0) the benchmark results you are experiencing in both 32 and 64 bit environments
1) the dataset that you're using which reproduces those benchmark results
2) instructions how to obtain those results