[QE-users] [QE-GPU] pDOS abruptly ends before Fermi Energy

Paolo Giannozzi paolo.giannozzi at uniud.it
Mon Jul 14 16:38:45 CEST 2025


Oh well. An input variable was not broadcast. Try this patch (with patch 
-p1):

diff --git a/PP/src/projwfc.f90 b/PP/src/projwfc.f90
index 9210fc3a5..88e2f7cef 100644
--- a/PP/src/projwfc.f90
+++ b/PP/src/projwfc.f90
@@ -125,6 +125,7 @@ PROGRAM do_projwfc
    CALL mp_bcast( lbinary_data,    ionode_id, intra_image_comm )
    CALL mp_bcast( lgww,      ionode_id, intra_image_comm )
    CALL mp_bcast( pawproj,   ionode_id, intra_image_comm )
+  CALL mp_bcast( kresolveddos,    ionode_id, intra_image_comm )
    CALL mp_bcast( tdosinboxes,     ionode_id, intra_image_comm )
    CALL mp_bcast( n_proj_boxes,    ionode_id, intra_image_comm )
    CALL mp_bcast( irmin,     ionode_id, intra_image_comm )


On 7/14/25 14:00, Paolo Giannozzi wrote:
> If it fits into a single GPU, run it with a single MPI process. In 
> parallel it is sometimes hard to figure out what fails how and why
> 
> PG
> 
> On 7/14/25 13:53, André Luiz Koch Liston wrote:
>>
>> Non ricevi spesso messaggi di posta elettronica da 
>> alk2244 at columbia.edu. Scopri perché è importante <https://aka.ms/ 
>> LearnAboutSenderIdentification>
>>
>>
>> Thanks, Stefano. Yes, the results are reproducible and consistent. 
>> Strangely, everything runs as expected when kresolveddos is set to 
>> false in {prefix}.projwfc.in <http://projwfc.in>. (Though that'd also 
>> defeat the very purpose of doing these calculations).
>> &PROJWFC
>> kresolveddos = .false.
>> ...
>> /
>>
>> I should also note that there's no difference whether the calculation 
>> preceding the "projwfc.x" step is a 'bands' or 'nscf' calculation.
>> I've even recompiled QE-7.2 (as per Vahid's suggestion, just in case) 
>> but the same error persists. I'm starting to run out of ideas.
>>
>> On Mon, Jul 14, 2025 at 4:04 AM Stefano de Gironcoli 
>> <degironc at sissa.it <mailto:degironc at sissa.it>> wrote:
>>
>>     __
>>     Thanks for your reply, I was asking because if things are working
>>     properly one would expect the sum of |psi|^2 should be less, but
>>     quite close, to the number of atomic wfs. and the sum of the pdos
>>     for each atomic wfc projection should sum to
>>     ZjQcmQRYFpfptBannerStart
>>     __
>>     This Message Is From an External Sender
>>     This message came from outside your organization.
>>     __
>>     ZjQcmQRYFpfptBannerEnd
>>
>>     Thanks for your reply,
>>
>>     I was asking because if things are working properly one would expect
>>     the sum of |psi|^2 should be less, but quite close, to the number of
>>     atomic wfs.
>>
>>     and the sum of the pdos for each atomic wfc projection should sum to
>>     (at most) 1
>>
>>     It seams that you are missing a lot of wfc weight (as you said in
>>     your original message). Something wrong in the generation of the
>>     atomic projectors maybe ...
>>
>>     Are your results reproducible ?
>>
>>     stefano
>>
>>     On 12/07/25 17:07, André Luiz Koch Liston wrote:
>>>     We're dealing with 1248 atomic orbitals (natomwfc = 1248) split
>>>     into Nb, Se, and Cl
>>>     Nb
>>>         Valence configuration:
>>>         nl pn  l   occ       Rcut    Rcut US       E pseu
>>>         4S  1  0  2.00      1.000      1.400    -4.288693
>>>         5S  2  0  1.00      1.000      1.400    -0.294951
>>>         4P  2  1  6.00      1.000      1.700    -2.535104
>>>         4D  3  2  4.00      1.000      1.700    -0.224572
>>>         Generation configuration:
>>>         4S  1  0  2.00      1.000      1.400    -4.288702
>>>         5S  2  0  1.00      1.000      1.400    -0.294957
>>>         4P  2  1  6.00      1.000      1.700    -2.535113
>>>         5P  3  1  0.00      1.000      1.700    -0.500000
>>>         4D  3  2  4.00      1.000      1.700    -0.224577
>>>         4D  3  2 -2.00      1.000      1.700     0.200000
>>>     Se
>>>     nl pn  l   occ               Rcut            Rcut US             E
>>>     pseu
>>>     4S  4  0  2.00      0.00000000000      1.50000000000     
>>> -1.27090753500
>>>     4P  4  1  4.00      0.00000000000      1.50000000000     
>>> -0.47850673200
>>>     Cl
>>>     nl pn  l   occ               Rcut            Rcut US             E
>>>     pseu
>>>     3S  3  0  2.00      0.00000000000      1.30000000000     
>>> -1.51543807600
>>>     3P  3  1  5.00      0.00000000000      1.30000000000     
>>> -0.62926841900
>>>
>>>     The sum of the projected |psi|^2 equals 556.315, well below
>>>     the 1800.00 electrons distributed along 1080 Kohn-Sham states from
>>>     the 'sfc' calculation (number of electrons = 1800.00, number of
>>>     Kohn-Sham states = 1080)
>>>
>>>
>>>     On Sat, Jul 12, 2025 at 3:44 AM Stefano de Gironcoli
>>>     <degironc at sissa.it <mailto:degironc at sissa.it>> wrote:
>>>
>>>         How many atomic orbitals do you have ? How much is the sum of
>>>         the projected |psi|^2 ? Stefano Get Outlook for Android From:
>>>         users <users-bounces@ lists. quantum-espresso. org>
>>>         <mailto:users-bounces@%E2%80%8Alists.%E2%80%8Aquantum-
>>>         espresso.%E2%80%8Aorg> on behalf of André Luiz Koch Liston
>>>         <alk2244@ columbia. edu> <mailto:alk2244@%E2%80%8Acolumbia.
>>>         %E2%80%8Aedu>
>>>         ZjQcmQRYFpfptBannerStart
>>>         This Message Is From an External Sender
>>>         This message came from outside your organization.
>>>         ZjQcmQRYFpfptBannerEnd
>>>         How many atomic orbitals do you have ?
>>>         How much is the sum of the projected |psi|^2 ?
>>>         Stefano
>>>
>>>         Get Outlook for Android <https://urldefense.proofpoint.com/v2/
>>>         url?
>>>         
>>> u=https-3A__aka.ms_AAb9ysg&d=DwMFoQ&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=-CyxoGv2Hs4gjxMybGsqVAOKuvXHk2d8dyOOhFrMP0M&m=5fokPCkla7G3j0EXdf60ebC1pD_eDNinDgcatpU-7tUL1sG-bN05GAZ2BcBsnOr-&s=pH2bm8bL5e4b_jqajXVLYxlrAkydn-aPtqW6B20HfcI&e=>
>>>         
>>> ------------------------------------------------------------------------
>>>         *From:* users <users-bounces at lists.quantum-espresso.org
>>>         <mailto:users-bounces at lists.quantum-espresso.org>> on behalf
>>>         of André Luiz Koch Liston <alk2244 at columbia.edu
>>>         <mailto:alk2244 at columbia.edu>>
>>>         *Sent:* Friday, July 11, 2025 11:25:06 PM
>>>         *To:* users at lists.quantum-espresso.org
>>>         <mailto:users at lists.quantum-espresso.org>
>>>         <users at lists.quantum-espresso.org <mailto:users at lists.quantum-
>>>         espresso.org>>
>>>         *Subject:* [QE-users] [QE-GPU] pDOS abruptly ends before Fermi
>>>         Energy
>>>         Dear QE users,
>>>         I'm working to get bands containing information on atomic
>>>         projections ("fat bands"). I am able to obtain a reasonable
>>>         band structure (Fermi energy = 6.9736 eV) and total
>>>         DOS. However, when projecting the results of the 'scf' and
>>>         'bands' calculations using projwfc.x, the .pdos files abruptly
>>>         end way before the Fermi energy is reached. Snippet of the
>>>         projwfc output, where |psi|^2 drops to zero:
>>>
>>>         ...
>>>         ==== e( 557) =     2.90731 eV ====
>>>              psi =
>>>         
>>> 0.010*[#1204]+0.010*[#1208]+0.010*[#1212]+0.010*[#1216]+0.010*[#1220]
>>>                   +0.010*[#1224]+...+0.001*[# 797]
>>>             |psi|^2 = 0.991
>>>         ==== e( 558) =     2.91135 eV ====
>>>              psi =
>>>         
>>> 0.011*[#1156]+0.011*[#1160]+0.011*[#1164]+0.011*[#1168]+0.009*[# 514]
>>>                   +0.009*[# 518]+0.009*[# 522]+..
>>>         +0.001*[#1004]+0.001*[#1008]+0.001*[#1012]
>>>                   +0.001*[#1016]+0.001*[#1020]+0.001*[#1024]
>>>             |psi|^2 = 0.987
>>>         ==== e( 559) =     2.91694 eV ====
>>>              psi =
>>>         
>>> 0.004*[#1156]+0.004*[#1164]+0.003*[#1160]+0.003*[#1168]+0.003*[#1004]
>>>                   +0.003*[#1020]+0.003*[#1239]+...
>>>         +0.001*[#1008]+0.001*[#1203]
>>>                   +0.001*[#1219]+0.001*[# 459]+0.001*[# 479]
>>>             |psi|^2 = 0.337
>>>         ==== e( 560) =     2.91818 eV ====
>>>
>>>             |psi|^2 = 0.000
>>>         ==== e( 561) =     2.92904 eV ====
>>>
>>>             |psi|^2 = 0.000
>>>         ==== e( 562) =     2.93008 eV ====
>>>         ...
>>>
>>>         As such, the outputs of {prefix}.pdos_atm#{(Cl)_wfc#1(s) also
>>>         abruptly end around 2.9 eV, way before the Fermi energy is
>>>         reached. Strangely, {prefix}.pdos_tot does contain DOS
>>>         reaching to the Fermi energy:
>>>
>>>         image.png
>>>         # ik    E (eV)  dos(E)    pdos(E)
>>>         ...
>>>             1    6.862  0.346E+02  0.000E+00
>>>             1    6.872  0.515E+02  0.000E+00
>>>             1    6.882  0.104E+02  0.000E+00
>>>             1    6.892  0.287E+00  0.000E+00
>>>             1    6.902  0.634E+00  0.000E+00
>>>             1    6.912  0.161E+02  0.000E+00
>>>             1    6.922  0.556E+02  0.000E+00
>>>             1    6.932  0.260E+02  0.000E+00
>>>             1    6.942  0.164E+01  0.000E+00
>>>             1    6.952  0.140E-01  0.000E+00
>>>             1    6.962  0.162E-04  0.000E+00
>>>         *    1    6.972  0.254E-08  0.000E+00              - ~Fermi
>>>         Energy*
>>>             1    6.982  0.538E-13  0.000E+00
>>>             1    6.992  0.000E+00  0.000E+00
>>>             1    7.002  0.000E+00  0.000E+00
>>>             1    7.012  0.000E+00  0.000E+00
>>>             1    7.022  0.236E-11  0.000E+00
>>>             1    7.032  0.574E-07  0.000E+00
>>>             1    7.042  0.189E-03  0.000E+00
>>>             1    7.052  0.845E-01  0.000E+00
>>>             1    7.062  0.510E+01  0.000E+00
>>>             1    7.072  0.417E+02  0.000E+00
>>>             1    7.082  0.461E+02  0.000E+00
>>>             1    7.092  0.690E+01  0.000E+00
>>>         ...
>>>
>>>         Code version: QE  7.4.1 (GPU enabled)
>>>         Pseudopotentials:
>>>          - Cl     35.453 cl_pbesol_v1.4.uspp.F.UPF
>>>          - Nb     92.90638 Nb.pbesol-spn-kjpaw_psl.0.3.0.UPF
>>>          - Se     78.96 se_pbesol_v1.uspp.F.UPF
>>>         nbnd = 1100
>>>         Calculation type: SCF (pw.x) → Bands (pw.x) → projwfc.x → 
>>> bands.x
>>>
>>>         Has this issue been seen before? What steps may I take to
>>>         troubleshoot this?
>>>
>>>         --         Thank you for your time,
>>>         André Koch Liston
>>>         PhD Researcher in Chemistry
>>>         Columbia University
>>>         
>>> _______________________________________________________________________________
>>>         The Quantum ESPRESSO Foundation stands in solidarity with all
>>>         civilians worldwide who are victims of terrorism, military
>>>         aggression, and indiscriminate warfare.
>>>         
>>> --------------------------------------------------------------------------------
>>>         Quantum ESPRESSO is supported by MaX (https://
>>>         urldefense.proofpoint.com/v2/url?
>>>         
>>> u=http-3A__www.max-2Dcentre.eu&d=DwICAg&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=-CyxoGv2Hs4gjxMybGsqVAOKuvXHk2d8dyOOhFrMP0M&m=5fokPCkla7G3j0EXdf60ebC1pD_eDNinDgcatpU-7tUL1sG-bN05GAZ2BcBsnOr-&s=YzsqTsqRs-_FHx5Ow3GyJ5pV1X5YiS3WkTB7YnHCS_Q&e= <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.max-2Dcentre.eu&d=DwICAg&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=-CyxoGv2Hs4gjxMybGsqVAOKuvXHk2d8dyOOhFrMP0M&m=5fokPCkla7G3j0EXdf60ebC1pD_eDNinDgcatpU-7tUL1sG-bN05GAZ2BcBsnOr-&s=YzsqTsqRs-_FHx5Ow3GyJ5pV1X5YiS3WkTB7YnHCS_Q&e=> )
>>>         users mailing list users at lists.quantum-espresso.org
>>>         <mailto:users at lists.quantum-espresso.org>
>>>         https://urldefense.proofpoint.com/v2/url?
>>>         
>>> u=https-3A__lists.quantum-2Despresso.org_mailman_listinfo_users&d=DwICAg&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=-CyxoGv2Hs4gjxMybGsqVAOKuvXHk2d8dyOOhFrMP0M&m=5fokPCkla7G3j0EXdf60ebC1pD_eDNinDgcatpU-7tUL1sG-bN05GAZ2BcBsnOr-&s=ZZqhiB5Wlt0dbI5WJCJBsQLRjGUrTlfUuU6C2VbJYe8&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.quantum-2Despresso.org_mailman_listinfo_users&d=DwICAg&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=-CyxoGv2Hs4gjxMybGsqVAOKuvXHk2d8dyOOhFrMP0M&m=5fokPCkla7G3j0EXdf60ebC1pD_eDNinDgcatpU-7tUL1sG-bN05GAZ2BcBsnOr-&s=ZZqhiB5Wlt0dbI5WJCJBsQLRjGUrTlfUuU6C2VbJYe8&e=>
>>>
>>>
>>>
>>>     --     Thank you for your time,
>>>     André Koch Liston
>>>     PhD Researcher in Chemistry
>>>
>>>     
>>> _______________________________________________________________________________
>>>     The Quantum ESPRESSO Foundation stands in solidarity with all 
>>> civilians worldwide who are victims of terrorism, military 
>>> aggression, and indiscriminate warfare.
>>>     
>>> --------------------------------------------------------------------------------
>>>     Quantum ESPRESSO is supported by MaX (www.max-centre.eu <https:// 
>>> urldefense.proofpoint.com/v2/url? 
>>> u=http-3A__www.max-2Dcentre.eu&d=DwMDaQ&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=-CyxoGv2Hs4gjxMybGsqVAOKuvXHk2d8dyOOhFrMP0M&m=0fIQgLafPNAGW_8az8DLNonwSbRGeH7YS-F29pOw8eExg5oio_3sJxr8MXnuccq0&s=BBq6-mHm-4VllGK6EYWdWzZb_ssoz_30SOHaXkpJVHI&e=>)
>>>     users mailing listusers at lists.quantum-espresso.org 
>>> <mailto:users at lists.quantum-espresso.org>
>>>     https://lists.quantum-espresso.org/mailman/listinfo/users 
>>> <https://urldefense.proofpoint.com/v2/url? 
>>> u=https-3A__lists.quantum-2Despresso.org_mailman_listinfo_users&d=DwMDaQ&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=-CyxoGv2Hs4gjxMybGsqVAOKuvXHk2d8dyOOhFrMP0M&m=0fIQgLafPNAGW_8az8DLNonwSbRGeH7YS-F29pOw8eExg5oio_3sJxr8MXnuccq0&s=e8Nr7ASRzEUAXSMdVSHjaugMc-Noti4iIMEo7YtIfqc&e=>
>>     
>> _______________________________________________________________________________
>>     The Quantum ESPRESSO Foundation stands in solidarity with all
>>     civilians worldwide who are victims of terrorism, military
>>     aggression, and indiscriminate warfare.
>>     
>> --------------------------------------------------------------------------------
>>     Quantum ESPRESSO is supported by MaX (https://
>>     urldefense.proofpoint.com/v2/url?
>>     
>> u=http-3A__www.max-2Dcentre.eu&d=DwICAg&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=-CyxoGv2Hs4gjxMybGsqVAOKuvXHk2d8dyOOhFrMP0M&m=0fIQgLafPNAGW_8az8DLNonwSbRGeH7YS-F29pOw8eExg5oio_3sJxr8MXnuccq0&s=BBq6-mHm-4VllGK6EYWdWzZb_ssoz_30SOHaXkpJVHI&e= <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.max-2Dcentre.eu&d=DwICAg&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=-CyxoGv2Hs4gjxMybGsqVAOKuvXHk2d8dyOOhFrMP0M&m=0fIQgLafPNAGW_8az8DLNonwSbRGeH7YS-F29pOw8eExg5oio_3sJxr8MXnuccq0&s=BBq6-mHm-4VllGK6EYWdWzZb_ssoz_30SOHaXkpJVHI&e=> )
>>     users mailing list users at lists.quantum-espresso.org
>>     <mailto:users at lists.quantum-espresso.org>
>>     https://urldefense.proofpoint.com/v2/url?
>>     
>> u=https-3A__lists.quantum-2Despresso.org_mailman_listinfo_users&d=DwICAg&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=-CyxoGv2Hs4gjxMybGsqVAOKuvXHk2d8dyOOhFrMP0M&m=0fIQgLafPNAGW_8az8DLNonwSbRGeH7YS-F29pOw8eExg5oio_3sJxr8MXnuccq0&s=e8Nr7ASRzEUAXSMdVSHjaugMc-Noti4iIMEo7YtIfqc&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.quantum-2Despresso.org_mailman_listinfo_users&d=DwICAg&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=-CyxoGv2Hs4gjxMybGsqVAOKuvXHk2d8dyOOhFrMP0M&m=0fIQgLafPNAGW_8az8DLNonwSbRGeH7YS-F29pOw8eExg5oio_3sJxr8MXnuccq0&s=e8Nr7ASRzEUAXSMdVSHjaugMc-Noti4iIMEo7YtIfqc&e=>
>>
>>
>>
>> -- 
>> Thank you for your time,
>> André Koch Liston
>> PhD Researcher in Chemistry
>>
>> _______________________________________________________________________________
>> The Quantum ESPRESSO Foundation stands in solidarity with all 
>> civilians worldwide who are victims of terrorism, military aggression, 
>> and indiscriminate warfare.
>> --------------------------------------------------------------------------------
>> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
>> users mailing list users at lists.quantum-espresso.org
>> https://lists.quantum-espresso.org/mailman/listinfo/users
> 

-- 
Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 206, 33100 Udine Italy, +39-0432-558216



More information about the users mailing list