Picture for Gu-Yeon Wei

Gu-Yeon Wei

S$^{3}$: Increasing GPU Utilization during Generative Inference for Higher Throughput

Add code
Jun 09, 2023
Viaarxiv icon

CAMEL: Co-Designing AI Models and Embedded DRAMs for Efficient On-Device Learning

Add code
May 04, 2023
Figure 1 for CAMEL: Co-Designing AI Models and Embedded DRAMs for Efficient On-Device Learning
Figure 2 for CAMEL: Co-Designing AI Models and Embedded DRAMs for Efficient On-Device Learning
Figure 3 for CAMEL: Co-Designing AI Models and Embedded DRAMs for Efficient On-Device Learning
Figure 4 for CAMEL: Co-Designing AI Models and Embedded DRAMs for Efficient On-Device Learning
Viaarxiv icon

MP-Rec: Hardware-Software Co-Design to Enable Multi-Path Recommendation

Add code
Feb 21, 2023
Viaarxiv icon

GPU-based Private Information Retrieval for On-Device Machine Learning Inference

Add code
Jan 27, 2023
Viaarxiv icon

PerfSAGE: Generalized Inference Performance Predictor for Arbitrary Deep Learning Models on Edge Devices

Add code
Jan 26, 2023
Viaarxiv icon

Architectural Implications of Embedding Dimension during GCN on CPU and GPU

Add code
Dec 01, 2022
Viaarxiv icon

OMU: A Probabilistic 3D Occupancy Mapping Accelerator for Real-time OctoMap at the Edge

Add code
May 06, 2022
Figure 1 for OMU: A Probabilistic 3D Occupancy Mapping Accelerator for Real-time OctoMap at the Edge
Figure 2 for OMU: A Probabilistic 3D Occupancy Mapping Accelerator for Real-time OctoMap at the Edge
Figure 3 for OMU: A Probabilistic 3D Occupancy Mapping Accelerator for Real-time OctoMap at the Edge
Figure 4 for OMU: A Probabilistic 3D Occupancy Mapping Accelerator for Real-time OctoMap at the Edge
Viaarxiv icon

Tabula: Efficiently Computing Nonlinear Activation Functions for Secure Neural Network Inference

Add code
Mar 05, 2022
Figure 1 for Tabula: Efficiently Computing Nonlinear Activation Functions for Secure Neural Network Inference
Figure 2 for Tabula: Efficiently Computing Nonlinear Activation Functions for Secure Neural Network Inference
Figure 3 for Tabula: Efficiently Computing Nonlinear Activation Functions for Secure Neural Network Inference
Figure 4 for Tabula: Efficiently Computing Nonlinear Activation Functions for Secure Neural Network Inference
Viaarxiv icon

Gradient Disaggregation: Breaking Privacy in Federated Learning by Reconstructing the User Participant Matrix

Add code
Jun 10, 2021
Figure 1 for Gradient Disaggregation: Breaking Privacy in Federated Learning by Reconstructing the User Participant Matrix
Figure 2 for Gradient Disaggregation: Breaking Privacy in Federated Learning by Reconstructing the User Participant Matrix
Figure 3 for Gradient Disaggregation: Breaking Privacy in Federated Learning by Reconstructing the User Participant Matrix
Figure 4 for Gradient Disaggregation: Breaking Privacy in Federated Learning by Reconstructing the User Participant Matrix
Viaarxiv icon

MAVFI: An End-to-End Fault Analysis Framework with Anomaly Detection and Recovery for Micro Aerial Vehicles

Add code
May 27, 2021
Figure 1 for MAVFI: An End-to-End Fault Analysis Framework with Anomaly Detection and Recovery for Micro Aerial Vehicles
Figure 2 for MAVFI: An End-to-End Fault Analysis Framework with Anomaly Detection and Recovery for Micro Aerial Vehicles
Figure 3 for MAVFI: An End-to-End Fault Analysis Framework with Anomaly Detection and Recovery for Micro Aerial Vehicles
Figure 4 for MAVFI: An End-to-End Fault Analysis Framework with Anomaly Detection and Recovery for Micro Aerial Vehicles
Viaarxiv icon