[aws-ofi-rccl] [package list]

aws-ofi-rccl/66b3b31-cpeGNU-22.08 (aws-ofi-rccl-66b3b31-cpeGNU-22.08.eb)

Install with the EasyBuild-user module:

eb aws-ofi-rccl-66b3b31-cpeGNU-22.08.eb -r

To access module help after installation and get reminded for which stacks and partitions the module is installed, use module spider aws-ofi-rccl/66b3b31-cpeGNU-22.08.

EasyConfig:

easyblock = 'ConfigureMake'

name = 'aws-ofi-rccl'
version = '66b3b31'

homepage = 'https://github.com/ROCmSoftwarePlatform/aws-ofi-rccl'

whatis = [
    'Description: AWS OFI RCCL is a plug-in which enables EC2 developers to use libfabric as a network provider.'
]

description = """
Machine learning frameworks running on top of AMD GPUs use a library called 
RCCL which provides standard collective communication routines for an arbitrary 
number of GPUs installed across single or multiple nodes.

This module implements a plug-in which maps RCCLs connection-oriented transport 
APIs to libfabric's connection-less reliable interface. This allows RCCL 
applications to take benefit of libfabric's transport layer services like 
reliable message support and operating system bypass.
"""

software_license_urls = ['https://github.com/ROCmSoftwarePlatform/aws-ofi-rccl/blob/cxi/LICENSE']

toolchain = {'name': 'cpeGNU', 'version': '22.08'}

sources = [{
    'filename': '%(name)s-%(version)s.tar.gz',
    'git_config': {
        'url': 'https://github.com/ROCmSoftwarePlatform',
        'repo_name': '%(name)s',
        'commit': '%(version)s',
    }
}]

dependencies = [
    ('rccl', '2.12.7')
]

preconfigopts = ' ./autogen.sh && '
configopts = (
    ' --with-libfabric=/opt/cray/libfabric/$(pkg-config --modversion libfabric) '
    ' --with-hip=$EBROOTROCM/hip '
    ' --with-rccl=$EBROOTRCCL '
)

sanity_check_paths = {
    'files': ['lib/librccl-net.so.0.0.0'],
    'dirs':  ['bin', 'lib']
}

moduleclass = 'devel'

[aws-ofi-rccl] [package list]