Visual C++ 2008 - good compiler/ linker settings for speed?

[silentshark]

The subject field says it all, really. What are good compiler flags to wring out good speed from the MS C++ 2008 compiler?

So far, I've found the following are good:

/O2 - optimise speed
/arch:SSE2 - compile for platforms supporting SSE2 instruction set

But there are probably some other ones I haven't come across?

(And feel free to tell me to go and use compiler X, if I really should be trying something else for high speed code these days)

Regards,
Tom

[Roman Hartmann]

You might want to add /GL

Roman

[silentshark]

You might want to add /GL

Roman

Ok, will do. Any other candidates? I've done some googling, but not found much - there's something about profile guided optimisation which I might have a gander at.

[Dann Corbit]

The subject field says it all, really. What are good compiler flags to wring out good speed from the MS C++ 2008 compiler?

So far, I've found the following are good:

/O2 - optimise speed
/arch:SSE2 - compile for platforms supporting SSE2 instruction set

But there are probably some other ones I haven't come across?

(And feel free to tell me to go and use compiler X, if I really should be trying something else for high speed code these days)

Regards,
Tom

The Intel compiler produces a little bit better results, most of the time.
Profile guided optimization produced a good gain almost all the time (for both Intel and MSVC++ compilers).

[vladstamate]

Hi,

The Intel compiler produces a little bit better results, most of the time.

Is that valid for AMD processors as well? Will it know to use some intrinsics like popcnt for example? I use then with the MSVC++ compiler.

Profile guided optimization produced a good gain almost all the time (for both Intel and MSVC++ compilers).

I can testify to that. I got roughly 12% increase in speed (both when doing perft as well as normal search) by using PGO. In Visual Studio that is done as simply as clicking 3 buttons and waiting your program to run for some time so it gets some performance data. I was quite impressed.

[Dann Corbit]

Hi,

The Intel compiler produces a little bit better results, most of the time.

Is that valid for AMD processors as well? Will it know to use some intrinsics like popcnt for example? I use then with the MSVC++ compiler.

Profile guided optimization produced a good gain almost all the time (for both Intel and MSVC++ compilers).

I can testify to that. I got roughly 12% increase in speed (both when doing perft as well as normal search) by using PGO. In Visual Studio that is done as simply as clicking 3 buttons and waiting your program to run for some time so it gets some performance data. I was quite impressed.

There are problems using the Intel toolset for AMD processors.

The most annoying of which is that the profiler castrates itself.

[Sven]

You might want to add /GL

Roman

Ok, will do. Any other candidates? I've done some googling, but not found much - there's something about profile guided optimisation which I might have a gander at.

/EH- /GR-

Sven

[rvida]

I can testify to that. I got roughly 12% increase in speed (both when doing perft as well as normal search) by using PGO. In Visual Studio that is done as simply as clicking 3 buttons and waiting your program to run for some time so it gets some performance data. I was quite impressed.

You are a lucky one. Or you just happen to write "branchy" code. In my experience with Critter there is only cca 5% speedup from PGO

[Dann Corbit]

I can testify to that. I got roughly 12% increase in speed (both when doing perft as well as normal search) by using PGO. In Visual Studio that is done as simply as clicking 3 buttons and waiting your program to run for some time so it gets some performance data. I was quite impressed.

You are a lucky one. Or you just happen to write "branchy" code. In my experience with Critter there is only cca 5% speedup from PGO

I have seen up to 30% for some engines.
I usually do the following sort of test during the profile sessions:
1. A couple hundred EPD test positions at 5 seconds each
2. At least ten games of engine verses engine.
3. Perft (if the engine has it) at depths 4/5/6
On the games, I have the GUI restart the engines after each game.

[alpha123]

I can testify to that. I got roughly 12% increase in speed (both when doing perft as well as normal search) by using PGO. In Visual Studio that is done as simply as clicking 3 buttons and waiting your program to run for some time so it gets some performance data. I was quite impressed.

You are a lucky one. Or you just happen to write "branchy" code. In my experience with Critter there is only cca 5% speedup from PGO

I suppose he could be an unlucky one and his non-PGO code is just slow....

Peter