Optimal Choice of Waveform for Library Characterization for Accurate Delay Calculation
Ajoy Mandal, Saili Shete
Abstract
Ajoy Mandal, Saili Shete
Abstract
Accuracy of waveform aware delay calculation approaches in STA tools require selection of an appropriate driver waveform during library characterization. The optimal waveform shape is dependent on process corner, voltage, temperature, parasitics and also the properties of the transistors involved. Traditionally used ramp waveform is not suitable for accuracy particularly at lower voltages and temperatures. Identification of waveform shape across a wide range of operating corners and for different transistor types can involve significant cost in terms of resources and time. This paper discusses an efficient approach for finding the driver waveform for library characterization. It also enlists the factors that influence the waveform shape and the related careabouts during the waveform identification process. Lastly, it proposes an approach to reduce the run-time and resources used to get the optimal characterization waveform at all operating corners.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Accuracy of waveform aware delay calculation approaches in STA tools require selection of an appropriate driver waveform during library characterization. The optimal waveform shape is dependent on process corner, voltage, temperature, parasitics and also the properties of the transistors involved. Traditionally used ramp waveform is not suitable for accuracy particularly at lower voltages and temperatures. Identification of waveform shape across a wide range of operating corners and for different transistor types can involve significant cost in terms of resources and time. This paper discusses an efficient approach for finding the driver waveform for library characterization. It also enlists the factors that influence the waveform shape and the related careabouts during the waveform identification process. Lastly, it proposes an approach to reduce the run-time and resources used to get the optimal characterization waveform at all operating corners.
Key concepts: Waveform, Identification (biology), Parasitic extraction, Computer science, Process (computing), Voltage, Electronic engineering, Engineering