Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: IEEE compliant floating point arithmetic (Read 2347 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

IEEE compliant floating point arithmetic

The Intel compiler has a switch which forces it to strictly obey the IEEE floating-point arithmetic standards. This stops the compiler from converting doubles to floats, divisions into multiplications, etc...

In the compilers help file Intel strongly recommend that this switch be used if floating-point arithmetic accuracy is important to the application (floating-point arithmetic accuracy is very important for Lame).

Has anyone tried compiling Lame with the Intel compiler using this switch? I would be very interested to see what the speed and size differences would be from a Microsoft compiler compiled version.

I don't have access to the Intel compiler currently otherwise I would have compiled it myself to see what the differences were. I also haven't used the Intel compiler in over 6 months so I don't remember what the switch was, but it shouldn't be hard to find in the help file.

Thanks,

bub.