Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Improve 2.0 dev mpi #112

Open
wants to merge 17 commits into
base: nest-gpu-2.0-dev-mpi
Choose a base branch
from

Conversation

golosio
Copy link
Collaborator

@golosio golosio commented Jan 28, 2024

Implemented abstract base class and derived template class for connections. Implemented two template specializations (12 bytes and 16 bytes). Reduced gpu memory overhead for remote connection creation. Improved MPI remote connection memory checker with automatic check and summary of the results.

…<ConnKeyT, ConnStructT> derived classes, with specialized template class ConnectionTemplate<conn12b_key, conn12b_struct> 12 byte connections
…ak CUDA memory usage of each MPI process, total CUDA memory available for all MPI processes, free CUDA memory available for all MPI processes. Adapted MPI connections CUDA memory check scripts to run on terminal and to check used CUDA memory against theoretical prediction automatically
…-tidy. Prepared wrappers for clang-format and clang-tidy compatibility with CUDA / CUB headers and scripts to run formatting and checks on all source c++/cuda files.
…o find automatically CUDA and MPI default header paths or to accept user-defined path lists for include files, CUDA or MPI headers
…les with clang-format and check them with clang-tidy
… connection memory check through script benchmark_terminal.sh. Writing some comments on connection-related code.
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This file should be removed

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

At line 39
Warning: number of bits representing node index is fixed "
"to 32 and cannot be modified with conn16b connection type"

But at line 127
max_node_nbits_ = 31;

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants