I am making an attempt to transform my program from 2020.4.311 to 2024.1 and I am getting fairly constant outcomes aside from the erfc() perform name when a small worth, 4.338867E-03, is assigned to the variable sampleZ within the following equation.  The previous model returns 0.4982690 for LOLP and the brand new model returns 0.4982691.   This causes vital variations in simulation outcomes on the finish.

LOLP = 1.0 – erfc(-sampleZ / sqrt(2.0)) * 0.5

Listed below are the compiling choices

/nologo /debug:minimal /O2 /Qparallel /assume:buffered_io /Qip /recursive /fpscomp:filesfromcmd /fpscomp:ioformat /fpscomp:logicals /fpscomp:normal /warn:interfaces /Qsave /fp:supply /module:”x64Release” /object:”x64Release” /Fd”x64Releasevc160.pdb” /traceback /test:bounds /libs:static /threads /winapp /Qmkl:parallel /c

Linker:

/OUT:”x64Releasehpcon.exe” /INCREMENTAL:NO /NOLOGO /MANIFEST:NO /SUBSYSTEM:CONSOLE /STACK:200000000 /IMPLIB:”H:LCGSRCIntelFortranHpcon_Trunkx64Releasehpcon.lib”

This system consists of IMSL libraries as proven within the screenshot beneath,

Additionally, I wrote a 10-line program with the erfc() perform name and handed the identical small worth, 4.338867E-03, utilizing the brand new compiler I bought again to anticipated worth, .4982690.

Let me know in the event you want any additional data.

Thanks,

Greenhills