Sounds like that one machine may have floating point problems. Intel has a CPU test suite http://download.intel.com/support/processors/sb/theintelprocessordiagnostict oolinwindowsuserguider.pdf
And IBM has a floating point test suite (single precision) which may help show if that one computer has a CPU problem. https://www.research.ibm.com/haifa/projects/verification/fpgen/ieeets.html And is that one problem machine overclocked by chance or is the CPU fan not working well (i.e. overheating)? And if anybody is willing to swap CPUs that would answer the CPU question too. 73 Mike W9MDB -----Original Message----- From: char...@sucklingfamily.free-online.co.uk [mailto:char...@sucklingfamily.free-online.co.uk] Sent: Saturday, June 20, 2015 3:06 PM To: wsjtdev Subject: [wsjt-devel] Different decodes on different PCs Hi All Has anyone any experience with same release of WSJT-X 1.6.0 having different correlation decodes from the same set of recorded files, please? We have come across an example where one PC decodes fewer periods than several others, and of the periods that are decoded on this "rogue" machine the quality number is lower. The same settings have been used on all machines, and the station with the odd result has two other PCs that decode with the same result that VK7MO and I get. This is happening with JT4F mode. 73 Charlie G3WDG ---------------------------------------------------------------------------- -- _______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/wsjt-devel ------------------------------------------------------------------------------ _______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/wsjt-devel
