.png){kind=icon}
.png){kind=icon}
Return
Dynamic Precision-Scalable Thermal Mapping Algorithm for Three Dimensional Systolic-Array Based Neural Network Accelerator
.png){kind=icon}
PRE
AI
DOI：10.1109/TCE.2024.3378706.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Nowadays, the systolic-array based accelerator has been used widely for the neural-network applications. Multiple systolic-array based accelerator chips can be stacked by the 3D IC technology to improve the performance of the neural-network applications. However, the 3D accelerator increases the power density and causes the overheating. To avoid the overheating, the sacrifice of the performance for the 3D accelerator under the thermal limitations is important. In this work, a dynamic precision-scalable thermal mapping algorithm (DPSTM) is proposed to change the active processing elements with different data precisions in the 3D accelerators dynamically. The goal is to control the power density and peak temperature of the 3D accelerator. Compared with the related works, DPSTM can reduce 29%-77% and 7%-73% latencies in AlexNet and ResNet-18 with 92-95 degrees C thermal limitations.
Keywords:
Three-dimensional displays
Power system measurements
Density measurement
Artificial neural networks
Temperature distribution
Stacking
Heuristic algorithms
Neural network
accelerator
thermal mapping
3D
Journal
IF:
10.9
Papers:
5.1K
Citations:
6.8K