A C + + program we run on an x86 platform twice as fast as on a comparable, if not faster Sparc platform On both platforms running Solaris 10 Can you help us with regard to the Compileroption or with an other tip there.

You may have to live with the difference.

Let me ask you, is the clock speed on the x86 roughly twice as fast as the clock speed of the SPARC platform? A lot of programs that are CPU intensive perform based on clock speed more than anything else. Even though the SPARC design is supposed to be "doing more" with fewer cycles, in real world applications I've found that you can predict performance by looking at the clock speed of the processors. You may be able to gain an advantage by using parallel processing or multiple threads, but a lot of times that requires a major rewrite of the program and that would benefit performance on either architecture.

This even extends to Java based apps. I've found that a Java JVM runs MUCH faster on Intel chips, regardless of the OS it's using. Again, this comes down to clock speed.

On the other hand, I would look at your compiler's options. Look for any SPARC specific parameters and then try them. You may find one that gives a nice boost. Also, try the various optimization options (-O2, -O3, etc). Again, it depends on what options your compiler has available.

Despite that, I don't think you're get them equal unless you find a SPARC chip that has a clock speed similar to the Intel chip you are using.