This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

g77 Help Wanted?


Dear Madam/Sir

Running a LAPACK's benchmark (1000d.f) on two machines:

1. PIII 450MHz, 128MB RAM
2. PII  300MHz, 128MB RAM

under Linux (Slachware 7, Kernel 2.2.13), with egcs-1.1.2.

I obtained the following results in terms of Mflops:

CASE 1:
Machine 1 (no compiler optimization options): 20Mflops
Machine 2 (no compiler optimization options): 12Mflops

CASE 2:
Machine 1 (f77 -O ) : 23 Mflops
Machine 2 (f77 -O) : 26.6 Mflops

Is this paradox (PII is faster!! than PIII in case 2) due to the fuct
that the curent fortan compiler
does not exploits the architecture of PIII?

Thank you in advance,
Antony Spyropoulos



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]