{ai}[foss/2024a] DP-GEN v0.13.2, DeePMD-kit v3.1.1, LAMMPS v29Aug2024_update2 w/ CUDA 12.6.0, kokkos deepmd CUDA 12.6.0 by JarneRenders · Pull Request #24733 · easybuilders/easybuild-easyconfigs

JarneRenders · 2025-12-02T08:56:52Z

(created using eb --new-pr)

Depends on:

…it-3.1.1-foss-2024a-CUDA-12.6.0.eb, LAMMPS-29Aug2024_update2-foss-2024a-kokkos-CUDA-12.6.0-with-deepmd-plugin.eb

github-actions · 2025-12-02T08:57:25Z

Updated software `LAMMPS-29Aug2024_update2-foss-2024a-kokkos-deepmd-CUDA-12.6.0.eb`

Diff against LAMMPS-2Aug2023_update2-foss-2024a-kokkos-CUDA-12.6.0.eb

easybuild/easyconfigs/l/LAMMPS/LAMMPS-2Aug2023_update2-foss-2024a-kokkos-CUDA-12.6.0.eb

diff --git a/easybuild/easyconfigs/l/LAMMPS/LAMMPS-2Aug2023_update2-foss-2024a-kokkos-CUDA-12.6.0.eb b/easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos-deepmd-CUDA-12.6.0.eb
index aa7ff10260..6e31150098 100644
--- a/easybuild/easyconfigs/l/LAMMPS/LAMMPS-2Aug2023_update2-foss-2024a-kokkos-CUDA-12.6.0.eb
+++ b/easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos-deepmd-CUDA-12.6.0.eb
@@ -1,7 +1,8 @@
 name = 'LAMMPS'
-version = '2Aug2023_update2'
+version = '29Aug2024_update2'
 _cuda_suffix = '-CUDA-%(cudaver)s'
-versionsuffix = '-kokkos' + _cuda_suffix
+versionsuffix = '-kokkos-deepmd' + _cuda_suffix
+_deepmdkitver = '3.1.1'
 
 homepage = 'https://www.lammps.org'
 description = """LAMMPS is a classical molecular dynamics code, and an acronym
@@ -18,25 +19,36 @@ functionality.
 toolchain = {'name': 'foss', 'version': '2024a'}
 toolchainopts = {'openmp': True, 'usempi': True, 'extra_cxxflags': '-D_AMXTILEINTRIN_H_INCLUDED'}
 # '-D_AMXTILEINTRIN_H_INCLUDED' flag is required to avoid the following errors,
-# likely due to an incompatibility between the GCC 13.3.0 and CUDA 12.6.0:
+# likely due to an incompatibility between the GCC 13.3.0 and NVCC CUDA 12.6.0:
 #   [...]/GCCcore/13.3.0/lib/gcc/x86_64-pc-linux-gnu/13.3.0/include/amxtileintrin.h(42):
 #     error: identifier "__builtin_ia32_ldtilecfg" is undefined
 #   [...]/GCCcore/13.3.0/lib/gcc/x86_64-pc-linux-gnu/13.3.0/include/amxtileintrin.h(49):
-#     error: identifier "__builtin_ia32_sttilecfg" is undefined
+# error: identifier "__builtin_ia32_sttilecfg" is undefined
 
 # 'https://github.com/lammps/lammps/archive/'
-source_urls = [GITHUB_LOWER_SOURCE]
-sources = ['stable_%(version)s.tar.gz']
+
+source_urls = [
+    GITHUB_LOWER_SOURCE,
+    'https://github.com/deepmodeling/deepmd-kit/archive/refs/tags/',
+]
+sources = [
+    'stable_%(version)s.tar.gz',
+    {
+        'download_filename': 'v%s.tar.gz' % _deepmdkitver,
+        'filename': 'deepmd-v%s.tar.gz' % _deepmdkitver,
+    },
+]
 patches = [
     'LAMMPS-2Aug2023_install_lammps_python_package_in_eb_software_module.patch',
-    'LAMMPS-2Aug2023_fix-timestep-balance-example.patch',
+    'LAMMPS-29Aug2024_install-deepmd-plugin.patch',
 ]
 checksums = [
-    {'stable_2Aug2023_update2.tar.gz': '3bcecabc9cad08d0a4e4d989b52d29c58505f7ead8ebacf43c9db8d9fd3d564a'},
+    {'stable_29Aug2024_update2.tar.gz': 'f8ca3f021a819ced8658055f7750e235c51b4937ddb621cf1bd7bee08e0b6266'},
+    {'deepmd-v3.1.1.tar.gz': 'f68f08db7c48479c0fc1f16b4dfd2e62c5dadcc229143ebd81bb802b6605d3bb'},
     {'LAMMPS-2Aug2023_install_lammps_python_package_in_eb_software_module.patch':
      '314ab1064af74ddfcf4091dfe9f90fa32cfa4100e8d01fea7b59b019575a7c51'},
-    {'LAMMPS-2Aug2023_fix-timestep-balance-example.patch':
-     '6f68387ced2b4bd0a06e4c3d31b0ffd47042476777d87d8a0ca6b19a4e6a1777'},
+    {'LAMMPS-29Aug2024_install-deepmd-plugin.patch':
+     'e601be73e131ed0740553b6ceff066eca0a7a474aad3d1b9e8dd4409ac6497df'},
 ]
 
 builddependencies = [
@@ -69,6 +81,7 @@ dependencies = [
     ('VTK', '9.3.1'),
     # We use a custom build of MDI
     ('MDI', '1.4.26'),
+    ('DeePMD-kit', _deepmdkitver, _cuda_suffix),
 ]
 if ARCH == 'x86_64':
     # TBB and ScaFaCos are an optional dependency when building on Intel arch
@@ -196,4 +209,6 @@ general_packages = [
 # note: only the highest capability will be used
 # cuda_compute_capabilities = ['9.0']
 
+configopts = '-Ddeepmd_source_dir=%%(builddir)s/deepmd-kit-%s' % _deepmdkitver
+
 moduleclass = 'chem'

Diff against LAMMPS-29Aug2024_update2-foss-2024a-kokkos.eb

easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos.eb

diff --git a/easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos.eb b/easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos-deepmd-CUDA-12.6.0.eb
index 669c4d81c5..6e31150098 100644
--- a/easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos.eb
+++ b/easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos-deepmd-CUDA-12.6.0.eb
@@ -1,6 +1,8 @@
 name = 'LAMMPS'
 version = '29Aug2024_update2'
-versionsuffix = '-kokkos'
+_cuda_suffix = '-CUDA-%(cudaver)s'
+versionsuffix = '-kokkos-deepmd' + _cuda_suffix
+_deepmdkitver = '3.1.1'
 
 homepage = 'https://www.lammps.org'
 description = """LAMMPS is a classical molecular dynamics code, and an acronym
@@ -15,18 +17,38 @@ functionality.
 """
 
 toolchain = {'name': 'foss', 'version': '2024a'}
-toolchainopts = {'openmp': True, 'usempi': True}
+toolchainopts = {'openmp': True, 'usempi': True, 'extra_cxxflags': '-D_AMXTILEINTRIN_H_INCLUDED'}
+# '-D_AMXTILEINTRIN_H_INCLUDED' flag is required to avoid the following errors,
+# likely due to an incompatibility between the GCC 13.3.0 and NVCC CUDA 12.6.0:
+#   [...]/GCCcore/13.3.0/lib/gcc/x86_64-pc-linux-gnu/13.3.0/include/amxtileintrin.h(42):
+#     error: identifier "__builtin_ia32_ldtilecfg" is undefined
+#   [...]/GCCcore/13.3.0/lib/gcc/x86_64-pc-linux-gnu/13.3.0/include/amxtileintrin.h(49):
+# error: identifier "__builtin_ia32_sttilecfg" is undefined
 
 # 'https://github.com/lammps/lammps/archive/'
-source_urls = [GITHUB_LOWER_SOURCE]
-sources = ['stable_%(version)s.tar.gz']
+
+source_urls = [
+    GITHUB_LOWER_SOURCE,
+    'https://github.com/deepmodeling/deepmd-kit/archive/refs/tags/',
+]
+sources = [
+    'stable_%(version)s.tar.gz',
+    {
+        'download_filename': 'v%s.tar.gz' % _deepmdkitver,
+        'filename': 'deepmd-v%s.tar.gz' % _deepmdkitver,
+    },
+]
 patches = [
     'LAMMPS-2Aug2023_install_lammps_python_package_in_eb_software_module.patch',
+    'LAMMPS-29Aug2024_install-deepmd-plugin.patch',
 ]
 checksums = [
     {'stable_29Aug2024_update2.tar.gz': 'f8ca3f021a819ced8658055f7750e235c51b4937ddb621cf1bd7bee08e0b6266'},
+    {'deepmd-v3.1.1.tar.gz': 'f68f08db7c48479c0fc1f16b4dfd2e62c5dadcc229143ebd81bb802b6605d3bb'},
     {'LAMMPS-2Aug2023_install_lammps_python_package_in_eb_software_module.patch':
      '314ab1064af74ddfcf4091dfe9f90fa32cfa4100e8d01fea7b59b019575a7c51'},
+    {'LAMMPS-29Aug2024_install-deepmd-plugin.patch':
+     'e601be73e131ed0740553b6ceff066eca0a7a474aad3d1b9e8dd4409ac6497df'},
 ]
 
 builddependencies = [
@@ -35,6 +57,9 @@ builddependencies = [
     ('archspec', '0.2.5'),
 ]
 dependencies = [
+    ('CUDA', '12.6.0', '', SYSTEM),
+    ('UCX-CUDA', '1.16.0', _cuda_suffix),
+    ('NCCL', '2.22.3', _cuda_suffix),
     ('Python', '3.12.3'),
     ('libpng', '1.6.43'),
     ('libjpeg-turbo', '3.0.1'),
@@ -56,6 +81,7 @@ dependencies = [
     ('VTK', '9.3.1'),
     # We use a custom build of MDI
     ('MDI', '1.4.26'),
+    ('DeePMD-kit', _deepmdkitver, _cuda_suffix),
 ]
 if ARCH == 'x86_64':
     # TBB and ScaFaCos are an optional dependency when building on Intel arch
@@ -179,4 +205,10 @@ general_packages = [
 # - MSCG
 # - QMMM (setup seems complex)
 
+# hardware-specific option
+# note: only the highest capability will be used
+# cuda_compute_capabilities = ['9.0']
+
+configopts = '-Ddeepmd_source_dir=%%(builddir)s/deepmd-kit-%s' % _deepmdkitver
+
 moduleclass = 'chem'

Diff against LAMMPS-29Aug2024_update2-foss-2024a-kokkos-CUDA-12.6.0.eb

easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos-CUDA-12.6.0.eb

diff --git a/easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos-CUDA-12.6.0.eb b/easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos-deepmd-CUDA-12.6.0.eb
index 11c82eb35a..6e31150098 100644
--- a/easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos-CUDA-12.6.0.eb
+++ b/easybuild/easyconfigs/l/LAMMPS/LAMMPS-29Aug2024_update2-foss-2024a-kokkos-deepmd-CUDA-12.6.0.eb
@@ -1,7 +1,8 @@
 name = 'LAMMPS'
 version = '29Aug2024_update2'
 _cuda_suffix = '-CUDA-%(cudaver)s'
-versionsuffix = '-kokkos' + _cuda_suffix
+versionsuffix = '-kokkos-deepmd' + _cuda_suffix
+_deepmdkitver = '3.1.1'
 
 homepage = 'https://www.lammps.org'
 description = """LAMMPS is a classical molecular dynamics code, and an acronym
@@ -25,15 +26,29 @@ toolchainopts = {'openmp': True, 'usempi': True, 'extra_cxxflags': '-D_AMXTILEIN
 # error: identifier "__builtin_ia32_sttilecfg" is undefined
 
 # 'https://github.com/lammps/lammps/archive/'
-source_urls = [GITHUB_LOWER_SOURCE]
-sources = ['stable_%(version)s.tar.gz']
+
+source_urls = [
+    GITHUB_LOWER_SOURCE,
+    'https://github.com/deepmodeling/deepmd-kit/archive/refs/tags/',
+]
+sources = [
+    'stable_%(version)s.tar.gz',
+    {
+        'download_filename': 'v%s.tar.gz' % _deepmdkitver,
+        'filename': 'deepmd-v%s.tar.gz' % _deepmdkitver,
+    },
+]
 patches = [
     'LAMMPS-2Aug2023_install_lammps_python_package_in_eb_software_module.patch',
+    'LAMMPS-29Aug2024_install-deepmd-plugin.patch',
 ]
 checksums = [
     {'stable_29Aug2024_update2.tar.gz': 'f8ca3f021a819ced8658055f7750e235c51b4937ddb621cf1bd7bee08e0b6266'},
+    {'deepmd-v3.1.1.tar.gz': 'f68f08db7c48479c0fc1f16b4dfd2e62c5dadcc229143ebd81bb802b6605d3bb'},
     {'LAMMPS-2Aug2023_install_lammps_python_package_in_eb_software_module.patch':
      '314ab1064af74ddfcf4091dfe9f90fa32cfa4100e8d01fea7b59b019575a7c51'},
+    {'LAMMPS-29Aug2024_install-deepmd-plugin.patch':
+     'e601be73e131ed0740553b6ceff066eca0a7a474aad3d1b9e8dd4409ac6497df'},
 ]
 
 builddependencies = [
@@ -66,6 +81,7 @@ dependencies = [
     ('VTK', '9.3.1'),
     # We use a custom build of MDI
     ('MDI', '1.4.26'),
+    ('DeePMD-kit', _deepmdkitver, _cuda_suffix),
 ]
 if ARCH == 'x86_64':
     # TBB and ScaFaCos are an optional dependency when building on Intel arch
@@ -193,4 +209,6 @@ general_packages = [
 # note: only the highest capability will be used
 # cuda_compute_capabilities = ['9.0']
 
+configopts = '-Ddeepmd_source_dir=%%(builddir)s/deepmd-kit-%s' % _deepmdkitver
+
 moduleclass = 'chem'

laraPPr · 2026-01-23T11:17:33Z

Another version of PyTorch for foss/2024a is now merged to you have a hard requirement for PyTorch/2.6.0 or could you use the newer version?

{ai}[foss/2024a] PyTorch v2.7.1 w/ CUDA 12.6.0 #23923

laraPPr · 2026-01-23T11:28:17Z

Oh your also stuck behind a pr of mine. That was one was waiting on a new feature in framework but I think it is now there. So hopefully that one will start moving as well.

JarneRenders · 2026-01-23T13:32:33Z

Another version of PyTorch for foss/2024a is now merged to you have a hard requirement for PyTorch/2.6.0 or could you use the newer version?

Newer version should not be an issue. I can have a look at this once I find some time.

adding easyconfigs: DP-GEN-0.13.2-foss-2024a-CUDA-12.6.0.eb, DeePMD-k…

a6014e3

…it-3.1.1-foss-2024a-CUDA-12.6.0.eb, LAMMPS-29Aug2024_update2-foss-2024a-kokkos-CUDA-12.6.0-with-deepmd-plugin.eb

github-actions bot added 2024a issues & PRs related to 2024a common toolchains new update labels Dec 2, 2025

JarneRenders and others added 2 commits December 2, 2025 15:06

add modloadmsg

d556f15

Make versionsuffix more concise

abd381f

JarneRenders changed the title ~~{ai}[foss/2024a] DP-GEN v0.13.2, DeePMD-kit v3.1.1, LAMMPS v29Aug2024_update2 w/ CUDA 12.6.0, kokkos CUDA 12.6.0 with deepmd plugin~~ {ai}[foss/2024a] DP-GEN v0.13.2, DeePMD-kit v3.1.1, LAMMPS v29Aug2024_update2 w/ CUDA 12.6.0, kokkos deepmd CUDA 12.6.0 Dec 4, 2025

laraPPr mentioned this pull request Jan 23, 2026

{ai}[foss/2023a] DeePMD-kit v3.0.1, Horovod v0.28.1 w/ TensorFlow 2.13.0 #22217

Open

laraPPr self-requested a review January 23, 2026 11:28

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

{ai}[foss/2024a] DP-GEN v0.13.2, DeePMD-kit v3.1.1, LAMMPS v29Aug2024_update2 w/ CUDA 12.6.0, kokkos deepmd CUDA 12.6.0#24733

{ai}[foss/2024a] DP-GEN v0.13.2, DeePMD-kit v3.1.1, LAMMPS v29Aug2024_update2 w/ CUDA 12.6.0, kokkos deepmd CUDA 12.6.0#24733
JarneRenders wants to merge 3 commits intoeasybuilders:developfrom
JarneRenders:20251202095644_new_pr_DP-GEN0132

JarneRenders commented Dec 2, 2025 •

edited by WilleBell

Loading

Uh oh!

github-actions bot commented Dec 2, 2025 •

edited

Loading

Uh oh!

laraPPr commented Jan 23, 2026

Uh oh!

laraPPr commented Jan 23, 2026

Uh oh!

JarneRenders commented Jan 23, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

JarneRenders commented Dec 2, 2025 • edited by WilleBell Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions bot commented Dec 2, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Updated software LAMMPS-29Aug2024_update2-foss-2024a-kokkos-deepmd-CUDA-12.6.0.eb

Uh oh!

laraPPr commented Jan 23, 2026

Uh oh!

laraPPr commented Jan 23, 2026

Uh oh!

JarneRenders commented Jan 23, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

JarneRenders commented Dec 2, 2025 •

edited by WilleBell

Loading

github-actions bot commented Dec 2, 2025 •

edited

Loading

Updated software `LAMMPS-29Aug2024_update2-foss-2024a-kokkos-deepmd-CUDA-12.6.0.eb`