Link Copied.
Method and Electronic Device for Nonlinear Function Approximation Using Static Random Access Memory (SRAM)-Based In-Memory Computing

中文版本

IDF1230.png

Opportunity

The most efficient Deep Neural Network (DNN) such as CNN and MLP IC designs benefit from In-memory Computing (IMC) approaches and have energy efficiency ~1000 TOPS/W. In contrast, all recent LSTM have yet to surpass the energy efficiency of 10 TOPS/W –two orders of magnitude less than the state-of-the-art CNN ICs. The major reason for this is that the published LSTM ICs use digital processing element (PE) based architecture. Our proposed method can break this efficiency wall by enabling implementation of nonlinear operators within memory and gain energy efficiency above 100TOPS/W for LSTM IC.

Technology 

The main function of this IP is to enable implementation of scalar or vector nonlinear functions within SRAM memory. In-memory computing enables a high energy efficiency over conventional digital implementation where data read from memory consumes a lot of energy. Scalar nonlinearities such as sigmoid, tanh and vector nonlinearities such as softmax are very commonly used in neural networks achieving state-of-the-art performance in various applications. 

Advantages

  • Low latency 
  • High energy efficiency 

Applications

  • Key-word spotting
  • Speech separation and speech enhancement
Remarks
IDF: 1230
IP Status
Patent filed
Technology Readiness Level (TRL)
6
Questions about this Technology?
Contact Our Tech Manager
Contact Our Tech Manager
Method and Electronic Device for Nonlinear Function Approximation Using Static Random Access Memory (SRAM)-Based In-Memory Computing

Personal Information

(ReCaptcha V3 Hidden Field)

We use cookies to ensure you get the best experience on our website.

More Information