-
Notifications
You must be signed in to change notification settings - Fork 45
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
Implementation of CUDA accelerated passive crossbar programming simulation for the 2021 Data Driven model #125
Commits on Nov 4, 2021
-
random_crossbar_init implemented
The following variable: random_crossbar_init initializes the crossbar to random device conductances in between 1/Ron and 1/Roff. This will be useful to test the robustness of a passive crossbar to reprogram itself to different conductances after previous inference. It will not affect anything if transistor is set to True.
Configuration menu - View commit details
-
Copy full SHA for 83f4e95 - Browse repository at this point
Copy the full SHA 83f4e95View commit details -
Configuration menu - View commit details
-
Copy full SHA for ddb86ce - Browse repository at this point
Copy the full SHA ddb86ceView commit details
Commits on Nov 11, 2021
-
Merge remote-tracking branch 'origin/master'
# Conflicts: # memtorch/bh/crossbar/Crossbar.py # memtorch/mn/Module.py
Configuration menu - View commit details
-
Copy full SHA for 6f4e50a - Browse repository at this point
Copy the full SHA 6f4e50aView commit details
Commits on Nov 26, 2021
-
Simulate_passive_optimizationV0.1
not tested not working just backup for now
Configuration menu - View commit details
-
Copy full SHA for 6f9da4d - Browse repository at this point
Copy the full SHA 6f9da4dView commit details
Commits on Dec 7, 2021
-
Configuration menu - View commit details
-
Copy full SHA for fa6479f - Browse repository at this point
Copy the full SHA fa6479fView commit details
Commits on Dec 21, 2021
-
Configuration menu - View commit details
-
Copy full SHA for 5e43d4d - Browse repository at this point
Copy the full SHA 5e43d4dView commit details
Commits on Jan 3, 2022
-
Data_Driven simulate no neighbours implemented implemented
Implementation of data_driven simulate no neighbours made.
Configuration menu - View commit details
-
Copy full SHA for 816a480 - Browse repository at this point
Copy the full SHA 816a480View commit details -
Configuration menu - View commit details
-
Copy full SHA for 8dec0e0 - Browse repository at this point
Copy the full SHA 8dec0e0View commit details
Commits on Jan 14, 2022
-
Initial Testing + Documentation
Shaky accuracy with simulate neighbours routine (probably working as intended, but the effects of V/2 applied is non neglectable) Documentation added for the cuda kernels
Configuration menu - View commit details
-
Copy full SHA for 1e20254 - Browse repository at this point
Copy the full SHA 1e20254View commit details
Commits on Feb 6, 2022
-
Adapted Crossbar.py to the current implementation of passive simulation
Changed the code of crossbar.py based on what is currently implemented in CUDA
Configuration menu - View commit details
-
Copy full SHA for 2b9bd1d - Browse repository at this point
Copy the full SHA 2b9bd1dView commit details -
Configuration menu - View commit details
-
Copy full SHA for e76c91b - Browse repository at this point
Copy the full SHA e76c91bView commit details -
Configuration menu - View commit details
-
Copy full SHA for cf4aa8f - Browse repository at this point
Copy the full SHA cf4aa8fView commit details -
random_crossbar_init implemented
The following variable: random_crossbar_init initializes the crossbar to random device conductances in between 1/Ron and 1/Roff. This will be useful to test the robustness of a passive crossbar to reprogram itself to different conductances after previous inference. It will not affect anything if transistor is set to True.
Configuration menu - View commit details
-
Copy full SHA for a3e8218 - Browse repository at this point
Copy the full SHA a3e8218View commit details -
Simulate_passive_optimizationV0.1
not tested not working just backup for now
Configuration menu - View commit details
-
Copy full SHA for cf4fed0 - Browse repository at this point
Copy the full SHA cf4fed0View commit details -
Configuration menu - View commit details
-
Copy full SHA for c6460c8 - Browse repository at this point
Copy the full SHA c6460c8View commit details -
Configuration menu - View commit details
-
Copy full SHA for 833877a - Browse repository at this point
Copy the full SHA 833877aView commit details -
Data_Driven simulate no neighbours implemented implemented
Implementation of data_driven simulate no neighbours made.
Configuration menu - View commit details
-
Copy full SHA for 9a46e56 - Browse repository at this point
Copy the full SHA 9a46e56View commit details -
Configuration menu - View commit details
-
Copy full SHA for 493fbe1 - Browse repository at this point
Copy the full SHA 493fbe1View commit details -
Initial Testing + Documentation
Shaky accuracy with simulate neighbours routine (probably working as intended, but the effects of V/2 applied is non neglectable) Documentation added for the cuda kernels
Configuration menu - View commit details
-
Copy full SHA for 3b5f552 - Browse repository at this point
Copy the full SHA 3b5f552View commit details -
Adapted Crossbar.py to the current implementation of passive simulation
Changed the code of crossbar.py based on what is currently implemented in CUDA
Configuration menu - View commit details
-
Copy full SHA for 432f147 - Browse repository at this point
Copy the full SHA 432f147View commit details -
Configuration menu - View commit details
-
Copy full SHA for 10ad7d3 - Browse repository at this point
Copy the full SHA 10ad7d3View commit details -
Configuration menu - View commit details
-
Copy full SHA for e1f8629 - Browse repository at this point
Copy the full SHA e1f8629View commit details -
Configuration menu - View commit details
-
Copy full SHA for beebe98 - Browse repository at this point
Copy the full SHA beebe98View commit details -
Configuration menu - View commit details
-
Copy full SHA for 55dbdb4 - Browse repository at this point
Copy the full SHA 55dbdb4View commit details -
Configuration menu - View commit details
-
Copy full SHA for 34a7862 - Browse repository at this point
Copy the full SHA 34a7862View commit details
Commits on Feb 7, 2022
-
fixed small bug in Crossbar.py
The code would have called memtorch_cuda_bindings even if the version was CPU.
Configuration menu - View commit details
-
Copy full SHA for de8dc2d - Browse repository at this point
Copy the full SHA de8dc2dView commit details
Commits on Feb 8, 2022
-
Update memtorch/cu/simulate_passive_kernels.cu
Co-authored-by: Corey Lammie <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 0db6b0c - Browse repository at this point
Copy the full SHA 0db6b0cView commit details -
Set CUDA = False for CI Co-authored-by: Corey Lammie <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 917f0ab - Browse repository at this point
Copy the full SHA 917f0abView commit details -
Update memtorch/bh/crossbar/Crossbar.py
remove memristor_model_params random comment oversight Co-authored-by: Corey Lammie <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 76c218d - Browse repository at this point
Copy the full SHA 76c218dView commit details -
Update memtorch/bh/crossbar/Crossbar.py
Co-authored-by: Corey Lammie <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for bb1b76a - Browse repository at this point
Copy the full SHA bb1b76aView commit details -
Co-authored-by: Corey Lammie <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 426e551 - Browse repository at this point
Copy the full SHA 426e551View commit details -
Update memtorch/cu/simulate_passive_kernels.cu
Co-authored-by: Corey Lammie <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 2233da0 - Browse repository at this point
Copy the full SHA 2233da0View commit details -
Update memtorch/cu/simulate_passive_kernels.cu
Co-authored-by: Corey Lammie <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 9b8c226 - Browse repository at this point
Copy the full SHA 9b8c226View commit details -
Update memtorch/cu/simulate_passive_kernels.cu
Co-authored-by: Corey Lammie <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 054e5dc - Browse repository at this point
Copy the full SHA 054e5dcView commit details -
Update based on recommendations
- Removed ideal_crossbar matrix - Made list of implemented models global - Added exception thrown when a model is not implemented - Refactored a part of the CUDA kernel -Removed function from CUD header (cuh file) -Removed redundant build_g_tensor function
Configuration menu - View commit details
-
Copy full SHA for cd7cd87 - Browse repository at this point
Copy the full SHA cd7cd87View commit details -
Configuration menu - View commit details
-
Copy full SHA for aab336b - Browse repository at this point
Copy the full SHA aab336bView commit details -
Configuration menu - View commit details
-
Copy full SHA for f08b1e1 - Browse repository at this point
Copy the full SHA f08b1e1View commit details -
Configuration menu - View commit details
-
Copy full SHA for e742b7b - Browse repository at this point
Copy the full SHA e742b7bView commit details