Fair enough I would definitely assume the 2026.1 version makes more sense than the 2024 version. For reference, I was interested in what the SMEAGOL interface would give, I translated (to the best of my ability) the regtest-negf-2 job into a set of jobs based on the examples given from "https://chemwiki.ch.ic.ac.uk/wiki/Potential_control_and_current_induced_forces_using_CP2K%2BSMEAGOL". I attached the results here. I may not have used SMEAGOL properly, as there are a number of available keywords for the CP2K+SMEAGOL interface that I didn't use, but the results are fairly close to CP2K internal routines. I would intuitively think the 2026 version of CP2K looks most reasonable between the 3: SMEAGOL, CP2K 2024 and CP2K 2026.<div><br /></div><div>I used the new version of CP2K 2026.1 on a previous job. The force/energy calculations are within numerical noise, but the fermi energy of the contacts and the shift in the HARTREE POTENTIAL were substantial.</div><div><br /></div><div>Version Ef Contact Shift in Hartree</div><div>2024.2 0.170 -0.011</div><div>2026.1 0.159 0.075</div><div><br /></div><div>Is this behavior consistent with the changes made to the CP2K NEGF routine (with regards to which properties changed more so than how they changed)?<br /><br /></div><div class="gmail_quote"><div dir="auto" class="gmail_attr">On Friday, January 23, 2026 at 10:57:45 PM UTC-8 Dmitry Ryndyk wrote:<br/></div><blockquote class="gmail_quote" style="margin: 0 0 0 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">Dear Michael,<div><br></div><div>One example. If you run regtest-negf-2, you may find the result in regtest-negf-2_2024.png, which is incorrect. New code gives regtest-negf-2_2024.png.</div><div><br></div><div>Best wishes,</div><div>Dmitry</div><div><br></div><div class="gmail_quote"><div dir="auto" class="gmail_attr">Michael LaCount schrieb am Samstag, 24. Januar 2026 um 06:28:23 UTC+1:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">I noticed in the 2026.1 release of CP2K a new feature 'NEGF: New method to extract the matrix Hamiltonians for electrodes'. I have been running NEGF calculations using an earlier version of CP2K 2024.2, is there a notable performance or accuracy difference with the new method (or from other changes to NEGF)?<br><br>I also just found the SMEAGOL NEGF interface option, I am unfamiliar with SMEAGOL and would like to ask if there is any advantage to using that interface vs the internal CP2K NEGF routines?</blockquote></div></blockquote></div>
<p></p>
-- <br />
You received this message because you are subscribed to the Google Groups "cp2k" group.<br />
To unsubscribe from this group and stop receiving emails from it, send an email to <a href="mailto:cp2k+unsubscribe@googlegroups.com">cp2k+unsubscribe@googlegroups.com</a>.<br />
To view this discussion visit <a href="https://groups.google.com/d/msgid/cp2k/cd43d1b3-3ec2-491a-b0c7-2bef01cbbcdan%40googlegroups.com?utm_medium=email&utm_source=footer">https://groups.google.com/d/msgid/cp2k/cd43d1b3-3ec2-491a-b0c7-2bef01cbbcdan%40googlegroups.com</a>.<br />