https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/PARDISO-produces-all-zero-result-randomly/m-p/866071#M7994 I found a workaround. After solving I check to see if the solution was modified, and if not, go and solve again. <DIV></DIV> <DIV></DIV> Mon, 22 Mar 2010 08:17:55 GMT crispybits 2010-03-22T08:17:55Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/PARDISO-produces-all-zero-result-randomly/m-p/866070#M7993 Hi. About once in every 1000 solves, I get all zeros as the solution. Any possible reason? It doesn't seem to depend on the input. If I simply run my program again it's fine. The error code is 0 when it happens. <DIV><SPAN style="font-family: Verdana, Arial, Helvetica, sans-serif;"><BR /></SPAN></DIV> <DIV><SPAN style="font-family: Verdana, Arial, Helvetica, sans-serif;">I just noticed that it isn't touching the solution vector. If I put a value in it before calling Pardiso, when it fails that value is still there.<BR /></SPAN> <DIV></DIV> <DIV></DIV> <DIV>MKL 10.2.2.025</DIV> <DIV>Real symmetric matrix</DIV> <DIV>Compiled with VC++ to a dll</DIV> <DIV>Called by VB.NET app</DIV> <DIV>In core</DIV> <DIV>Reordering type 2 (METIS)</DIV> <DIV>1 RHS vector</DIV> <DIV>2-cores</DIV> <DIV></DIV> <DIV></DIV> </DIV> Fri, 19 Mar 2010 12:28:03 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/PARDISO-produces-all-zero-result-randomly/m-p/866070#M7993 crispybits 2010-03-19T12:28:03Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/PARDISO-produces-all-zero-result-randomly/m-p/866071#M7994 I found a workaround. After solving I check to see if the solution was modified, and if not, go and solve again. <DIV></DIV> <DIV></DIV> Mon, 22 Mar 2010 08:17:55 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/PARDISO-produces-all-zero-result-randomly/m-p/866071#M7994 crispybits 2010-03-22T08:17:55Z