Re: Single vs. Double precision

From: Floris Buelens (floris_buelens_at_yahoo.com)
Date: Mon Jan 19 2009 - 09:58:06 CST

I was wondering about this as well - I noticed a compile time option (-DSHORTREALS) for single precision in common.h and tried it but never managed to make a working binary. And how about the new CUDA code in CVS, is that single or double precision? best regards, Floris ________________________________ From: Neelanjana Sengupta <senguptan_at_gmail.com> To: NAMD <namd-l_at_ks.uiuc.edu> Sent: Monday, 19 January, 2009 11:27:27 Subject: namd-l: Single vs. Double precision Hello NAMD community, I gather from a previous conversation that the NAMD code uses double precision. However, can the code be compiled as a single precision version? And if so, has anybody compared results and found serious deviations between the single and double precision results for any system? Some information would be appreciated. Thanks, Neelanjana -- ~~~~~~~~~~~~~~~~~~~~ Neelanjana Sengupta, PhD Physical and Materials Chemistry Division National Chemical Laboratory Dr. Homi Bhaba Road Pune 411008, India Phone: +91-20-2590 2087 ~~~~~~~~~~~~~~~~~~~~

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:52:15 CST