[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ccp4bb]: Benchmarks

***  For details on how to be removed from this list visit the  ***
***    CCP4 home page http://www.dl.ac.uk/CCP/CCP4/main.html    ***

Very interesting. My own tests using ccp4 software (mlphare, fft,
refmac) on similar systems to your put an Athlon 800 about 2.5x faster
than a MIPS R10000/195.

Possible explanation include:

1. Different compilation options. ccp4 is compiled -O2 on SGI, and -O1
on Linux. If solve were to use -O3 on SGI or -O0 on linux that would
explain the difference.

2. Different memory usage. SGIs have very good memory bandwidth, PC's
are a bit behind and depend on the type of memory and motherboard. Very
intensive random searching of large data arrays could explain it.

3. Different calculation type. The ccp4 programs I used mainly involve
single precision floating point and fairly intensive integer calcs for
array indexing. Double precision will tend to cripple x86 relative to