From: Mudrovcic, Korana <Korana.Mudrovcic.uni-duesseldorf.de>
Date: Tue, 9 Feb 2021 14:22:56 +0000

Dear all,

In the AMBER archive I found a couple of mentions of the floating-point exceptions flags signalling during the simulations, namely that the IEEE_UNDERFLOW_FLAG IEEE_DENORMAL are not of concern. When it came to IEEE_OVERFLOW_FLAG however, some responses were to check the system for clashes. In the (to my knowledge) last post related to the topic, it was suggested that the "garbage values" from previous calculations were causing the flag due to one of the variables not being initialized during setup and that the calculation results were not affected (http://dev-archive.ambermd.org/201811/0003.html).

In our research group, our calculations get the following note:

Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_OVERFLOW_FLAG IEEE_UNDERFLOW_FLAG IEEE_DENORMAL

This note is signalling with AMBER20 as well as with AMBER18 (we are using CUDA 10.2 version) and is also triggered in the tests despite all of them passing. This makes us believe that the results are not affected, and that the warning is harmless (?).

Are there any new insights about this note? The first instinct of my research colleagues was also to suspect clashes in the system for those who noticed the flag in their calculations, similar to the responses that were received to the question in the AMBER list. If the note is indeed harmless and its occurrence is widespread, maybe it would provide some peace of mind if it was officially mentioned somewhere that this is to be expected.

Thank you in advance.

Best regards,



Korana Mudrovcic

PhD student

Heinrich-Heine-Universität Düsseldorf

Institut für Pharmazeutische und Medizinische Chemie

Universitätsstraße 1 / Gebäude 26.23.

40225 Düsseldorf

Tel. : 0211 81-13830

Room :

AMBER mailing list
Received on Tue Feb 09 2021 - 06:30:02 PST
Custom Search