[aws] AWS OFI NCCL v1.1.0

This release requires Libfabric v1.11.0 and supports NCCL v2.7.8 while maintaining backward compatibility with older NCCL versions (upto NCCL v2.4.8). It adds multi-NIC environment and GPUDirect RDMA support.

The plugin has been tested with following libfabric providers using unit tests bundled in the source code:

• efa
• tcp;ofi_rxm
• sockets

AWS OFI NCCL v1.1.0

This release requires Libfabric v1.11.0 and supports NCCL v2.7.8 while maintaining backward compatibility with older NCCL versions (upto NCCL v2.4.8). It adds multi-NIC environment and GPUDirect RDMA support.

The plugin has been tested with following libfabric providers using unit tests bundled in the source code:

• efa
• tcp;ofi_rxm
• sockets

[aws] AWS OFI NCCL v1.0.1

This release supports NCCL v2.6.4 while maintaining backward compatibility with older NCCL versions (upto NCCL v2.4.8). It also includes bug fixes and testing enhancements.

The plugin has been tested with following libfabric providers using unit tests bundled in the source code:

• efa
• tcp;ofi_rxm
• sockets

AWS OFI NCCL v1.0.1

This release supports NCCL v2.6.4 while maintaining backward compatibility with older NCCL versions (upto NCCL v2.4.8). It also includes bug fixes and testing enhancements.

The plugin has been tested with following libfabric providers using unit tests bundled in the source code:

• efa
• tcp;ofi_rxm
• sockets

AWS OFI NCCL v1.0.0

This release requires Libfabric v1.9.x and supports NCCL v2.5.6. It introduces changes to remove FI_AV_TABLE requirement from libfabric providers and provide several bug fixes including fixing overflow issues, memory leaks and adding completion checks for connection establishment APIs.

The plugin has been tested with following libfabric providers using unit tests bundled in the source code:

• tcp;ofi_rxm
• sockets
• efa

AWS OFI NCCL v0.9.2

This release introduces changes required to support NCCLv2.4 and fixes race
condition during connection establishment by removing FI_SOURCE requirement.

The plugin has been tested with following libfabric providers:

• tcp;ofi_rxm
• sockets
• verbs;ofi_rxm

AWS OFI NCCL v0.9.1

This release makes improvements to the building and CI infrastructure. It also
includes several bug fixes.

The plugin has been tested with following libfabric providers:

• tcp;ofi_rxm
• sockets
• verbs;ofi_rxm
• psm2

AWS OFI NCCL v0.9

AWS OFI NCCL supports NCCL v2.3.7+ and requires libfabric v1.6.x+.
Please note that current master of libfabric is broken for rxm providers and would require PR-4641.

The plugin has been tested with following libfabric providers:

• tcp;ofi_rxm
• sockets
• verbs;ofi_rxm