Picture for Sourjya Roy

Sourjya Roy

Evaluation of STT-MRAM as a Scratchpad for Training in ML Accelerators

Add code
Aug 03, 2023
Viaarxiv icon

PIM-DRAM: Accelerating Machine Learning Workloads using Processing in Commodity DRAM

Add code
May 14, 2021
Figure 1 for PIM-DRAM: Accelerating Machine Learning Workloads using Processing in Commodity DRAM
Figure 2 for PIM-DRAM: Accelerating Machine Learning Workloads using Processing in Commodity DRAM
Figure 3 for PIM-DRAM: Accelerating Machine Learning Workloads using Processing in Commodity DRAM
Figure 4 for PIM-DRAM: Accelerating Machine Learning Workloads using Processing in Commodity DRAM
Viaarxiv icon

Pruning Filters while Training for Efficiently Optimizing Deep Learning Networks

Add code
Mar 05, 2020
Figure 1 for Pruning Filters while Training for Efficiently Optimizing Deep Learning Networks
Figure 2 for Pruning Filters while Training for Efficiently Optimizing Deep Learning Networks
Figure 3 for Pruning Filters while Training for Efficiently Optimizing Deep Learning Networks
Figure 4 for Pruning Filters while Training for Efficiently Optimizing Deep Learning Networks
Viaarxiv icon

TxSim:Modeling Training of Deep Neural Networks on Resistive Crossbar Systems

Add code
Feb 25, 2020
Figure 1 for TxSim:Modeling Training of Deep Neural Networks on Resistive Crossbar Systems
Figure 2 for TxSim:Modeling Training of Deep Neural Networks on Resistive Crossbar Systems
Figure 3 for TxSim:Modeling Training of Deep Neural Networks on Resistive Crossbar Systems
Figure 4 for TxSim:Modeling Training of Deep Neural Networks on Resistive Crossbar Systems
Viaarxiv icon

Gradual Channel Pruning while Training using Feature Relevance Scores for Convolutional Neural Networks

Add code
Feb 23, 2020
Figure 1 for Gradual Channel Pruning while Training using Feature Relevance Scores for Convolutional Neural Networks
Figure 2 for Gradual Channel Pruning while Training using Feature Relevance Scores for Convolutional Neural Networks
Figure 3 for Gradual Channel Pruning while Training using Feature Relevance Scores for Convolutional Neural Networks
Figure 4 for Gradual Channel Pruning while Training using Feature Relevance Scores for Convolutional Neural Networks
Viaarxiv icon

Proposal for a Leaky-Integrate-Fire Spiking Neuron based on Magneto-Electric Switching of Ferro-magnets

Add code
Sep 29, 2016
Figure 1 for Proposal for a Leaky-Integrate-Fire Spiking Neuron based on Magneto-Electric Switching of Ferro-magnets
Figure 2 for Proposal for a Leaky-Integrate-Fire Spiking Neuron based on Magneto-Electric Switching of Ferro-magnets
Figure 3 for Proposal for a Leaky-Integrate-Fire Spiking Neuron based on Magneto-Electric Switching of Ferro-magnets
Figure 4 for Proposal for a Leaky-Integrate-Fire Spiking Neuron based on Magneto-Electric Switching of Ferro-magnets
Viaarxiv icon