<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<style type="text/css" style="display:none;"> P {margin-top:0;margin-bottom:0;} </style>
</head>
<body dir="ltr">
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
Hello</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
you don't need to specify anything special for running with GPUs; the input and options are the same, </div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
the first thing to check is whether you are using the right mpirun wrapper, the path returned by which mpirun should be something like this</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
also check from the output of nvidia-smi command if your driver and cuda versions are compatible (see here
<a href="https://docs.nvidia.com/deploy/cuda-compatibility/" id="LPlnk259190">https://docs.nvidia.com/deploy/cuda-compatibility/</a> )</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
if this doesn't help, please send more information</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
Pietro</div>
<div id="appendonsend"></div>
<div style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
<br>
</div>
<hr style="display: inline-block; width: 98%;">
<div dir="ltr" id="divRplyFwdMsg"><span style="font-family: Calibri, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);"><b>From:</b> R T Phillips <rtp1@cam.ac.uk><br>
<b>Sent:</b> Thursday, January 30, 2025 14:18<br>
<b>To:</b> Pietro Davide Delugas <pdelugas@sissa.it>; users@lists.quantum-espresso.org <users@lists.quantum-espresso.org>; paolo.giannozzi@uniud.it <paolo.giannozzi@uniud.it><br>
<b>Subject:</b> Re: [QE-users] QE-GPU - cudafor not found</span>
<div> </div>
</div>
<p style="margin-top: 0px; margin-bottom: 0px;">Dear Pietro, and also Paolo</p>
<p style="margin-top: 0px; margin-bottom: 0px;">Many thanks for the swift response. </p>
<p style="margin-top: 0px; margin-bottom: 0px;">--with-cuda-runtime=12.6 ... this leads to success in config.</p>
<p style="margin-top: 0px; margin-bottom: 0px;">make all seemed to complete successfully.</p>
<p style="margin-top: 0px; margin-bottom: 0px;">Running pw with mpirun ... so far no good.</p>
<p style="margin-top: 0px; margin-bottom: 0px;">I would be very grateful if you have any suggestion for the right options to select to run pw in the present configuration. I seem to have omitted specifying an hca device, and have hcoll errors.</p>
<p style="margin-top: 0px; margin-bottom: 0px;">Many thanks, I really appreciate your help.</p>
<p style="margin-top: 0px; margin-bottom: 0px;">Richard</p>
<p style="margin-top: 0px; margin-bottom: 0px;"><br>
</p>
<p style="margin-top: 0px; margin-bottom: 0px;"><br>
</p>
<div>On 30/01/2025 11:33, Pietro Davide Delugas wrote:</div>
<blockquote>
<div style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
Hello</div>
<div style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
you need to specifu --with-cuda-runtime=<value>, where value is a version of the runtime supported by your compiler, I guess 11.8 should do.</div>
<div style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
Hope it helps</div>
<div style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
Pietro</div>
<div style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
<br>
</div>
<hr style="display: inline-block; width: 98%;">
<div id="x_divRplyFwdMsg" dir="ltr"><span style="font-family: Calibri, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);"><b>From:</b> users
<a href="mailto:users-bounces@lists.quantum-espresso.org" id="OWA5fd96f1e-8a1d-abc4-21f0-156043ff83b3" class="x_moz-txt-link-rfc2396E OWAAutoLink">
<users-bounces@lists.quantum-espresso.org></a> on behalf of R T Phillips via users
<a href="mailto:users@lists.quantum-espresso.org" id="OWA87b71541-810f-6f1d-a3b7-6cb0332099ef" class="x_moz-txt-link-rfc2396E OWAAutoLink">
<users@lists.quantum-espresso.org></a><br>
<b>Sent:</b> Thursday, January 30, 2025 12:23<br>
<b>To:</b> <a href="mailto:users@lists.quantum-espresso.org" id="OWA4e64ce28-b315-b36e-ed38-369afe52fa91" class="x_moz-txt-link-abbreviated OWAAutoLink">users@lists.quantum-espresso.org</a>
<a href="mailto:users@lists.quantum-espresso.org" id="OWAb256bfa2-b6cd-d736-8f0f-a28a53183d1e" class="x_moz-txt-link-rfc2396E OWAAutoLink">
<users@lists.quantum-espresso.org></a><br>
<b>Subject:</b> [QE-users] QE-GPU - cudafor not found</span>
<div> </div>
</div>
<div style="font-size: 11pt;">I have been trying to compile Quantum Espresso for use on this system:<br>
Intel i9-14900 (24 core)<br>
Linux Mint 22.1 Cinnamon version 6.4.6<br>
Nvidia GeForce RTX 4070<br>
<br>
with:<br>
nvhpc_2024_2411_Linux_x86_64_cuda_12.6<br>
<br>
openmpi5.0.6<br>
<br>
qe 7.3.1<br>
<br>
PATH and LD_LIBRARY_PATH point to all the recommended folders.<br>
<br>
Compilers set to nvc, nvc++, nvfortran from nvhpc; mpif90 from openmpi<br>
also tried.<br>
<br>
./configure --enable-parallel --with-cuda=$NVHPC_CUDA_HOME<br>
--with-cuda-cc=89 --with-cuda-mpi=yes --enable-openmp<br>
<br>
this stops with:<br>
<br>
checking whether Fortran compiler accepts -cuda -gpu=cuda10.1... no<br>
configure: error: You do not have the cudafor module. Are you using<br>
NVHPC compiler?<br>
<br>
... well, yes. What then?<br>
<br>
I have searched for info about whether the nvhpc should include cudafor,<br>
and if so, how to make it visible. There are discussions related to<br>
distributions prior to hpc_sdk, but I can't find up-to-date advice ie<br>
after cuda was incorporated in the hpc_sdk. I switched to F90=mpif90<br>
from openmpi-5.0.6 and hit the same problem.<br>
Is there a solution to this? Clearly I want to use all the cores in the<br>
CPU and those in the GPU to optimise throughput.<br>
Thanks to anyone who can help!<br>
<br>
Richard Phillips, Cavendish Laboratory, UK<br>
<br>
_______________________________________________________________________________<br>
The Quantum ESPRESSO Foundation stands in solidarity with all civilians worldwide who are victims of terrorism, military aggression, and indiscriminate warfare.<br>
--------------------------------------------------------------------------------<br>
Quantum ESPRESSO is supported by MaX (<a href="http://www.max-centre.eu" id="OWA4b33769d-e935-03ad-83b6-86cffb68bd48" class="x_OWAAutoLink" data-auth="NotApplicable">www.max-centre.eu</a>)<br>
users mailing list <a href="mailto:users@lists.quantum-espresso.org" id="OWAf39f2d8e-25c8-84fe-4a26-91c74b4f5588" class="x_moz-txt-link-abbreviated OWAAutoLink">
users@lists.quantum-espresso.org</a><br>
<a href="https://lists.quantum-espresso.org/mailman/listinfo/users" id="OWAd4843aa5-63ec-9ec9-2396-d53c8d4b8273" class="x_OWAAutoLink x_moz-txt-link-freetext" data-auth="NotApplicable">https://lists.quantum-espresso.org/mailman/listinfo/users</a></div>
<div style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
<br>
</div>
<hr style="display: inline-block; width: 98%;">
<div id="x_divRplyFwdMsg" dir="ltr"><span style="font-family: Calibri, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);"><b>From:</b> users
<a href="mailto:users-bounces@lists.quantum-espresso.org" id="OWAf81527c4-44f4-2407-9464-cce7407c3eee" class="x_moz-txt-link-rfc2396E OWAAutoLink">
<users-bounces@lists.quantum-espresso.org></a> on behalf of R T Phillips via users
<a href="mailto:users@lists.quantum-espresso.org" id="OWAca0dcc91-b79c-7e33-fc6b-afb349403fde" class="x_moz-txt-link-rfc2396E OWAAutoLink">
<users@lists.quantum-espresso.org></a><br>
<b>Sent:</b> Thursday, January 30, 2025 12:23<br>
<b>To:</b> <a href="mailto:users@lists.quantum-espresso.org" id="OWAa101c7f0-f87c-04b7-ed20-7a28d0bfd6f9" class="x_moz-txt-link-abbreviated OWAAutoLink">users@lists.quantum-espresso.org</a>
<a href="mailto:users@lists.quantum-espresso.org" id="OWA99a150c3-08e2-2ff0-02d1-f216e5fcfcc6" class="x_moz-txt-link-rfc2396E OWAAutoLink">
<users@lists.quantum-espresso.org></a><br>
<b>Subject:</b> [QE-users] QE-GPU - cudafor not found</span>
<div> </div>
</div>
<div style="font-size: 11pt;">I have been trying to compile Quantum Espresso for use on this system:<br>
Intel i9-14900 (24 core)<br>
Linux Mint 22.1 Cinnamon version 6.4.6<br>
Nvidia GeForce RTX 4070<br>
<br>
with:<br>
nvhpc_2024_2411_Linux_x86_64_cuda_12.6<br>
<br>
openmpi5.0.6<br>
<br>
qe 7.3.1<br>
<br>
PATH and LD_LIBRARY_PATH point to all the recommended folders.<br>
<br>
Compilers set to nvc, nvc++, nvfortran from nvhpc; mpif90 from openmpi<br>
also tried.<br>
<br>
./configure --enable-parallel --with-cuda=$NVHPC_CUDA_HOME<br>
--with-cuda-cc=89 --with-cuda-mpi=yes --enable-openmp<br>
<br>
this stops with:<br>
<br>
checking whether Fortran compiler accepts -cuda -gpu=cuda10.1... no<br>
configure: error: You do not have the cudafor module. Are you using<br>
NVHPC compiler?<br>
<br>
... well, yes. What then?<br>
<br>
I have searched for info about whether the nvhpc should include cudafor,<br>
and if so, how to make it visible. There are discussions related to<br>
distributions prior to hpc_sdk, but I can't find up-to-date advice ie<br>
after cuda was incorporated in the hpc_sdk. I switched to F90=mpif90<br>
from openmpi-5.0.6 and hit the same problem.<br>
Is there a solution to this? Clearly I want to use all the cores in the<br>
CPU and those in the GPU to optimise throughput.<br>
Thanks to anyone who can help!<br>
<br>
Richard Phillips, Cavendish Laboratory, UK<br>
<br>
_______________________________________________________________________________<br>
The Quantum ESPRESSO Foundation stands in solidarity with all civilians worldwide who are victims of terrorism, military aggression, and indiscriminate warfare.<br>
--------------------------------------------------------------------------------<br>
Quantum ESPRESSO is supported by MaX (<a href="http://www.max-centre.eu" id="OWA7eaf3ddc-3789-0465-14da-ce58df3e604d" class="x_OWAAutoLink" data-auth="NotApplicable">www.max-centre.eu</a>)<br>
users mailing list <a href="mailto:users@lists.quantum-espresso.org" id="OWAf150ee4c-c390-5f4e-e67c-ddbfef9ca5ac" class="x_moz-txt-link-abbreviated OWAAutoLink">
users@lists.quantum-espresso.org</a><br>
<a href="https://lists.quantum-espresso.org/mailman/listinfo/users" id="OWA73f30275-ee86-ad61-ee30-f8037692a160" class="x_OWAAutoLink x_moz-txt-link-freetext" data-auth="NotApplicable">https://lists.quantum-espresso.org/mailman/listinfo/users</a></div>
</blockquote>
</body>
</html>