Growing digitization of integrated circuits has contributed to making system–on–chips ever more complex. Yet, it does not prevent that a substantial portion of a chip consists of analog and mixed–signal (AMS) circuits that provide critical functionality like signal conversion. Aggressive scaling of IC technologies, as well as advancing the integration of heterogeneous physical domains on chip, substantially complicates the design of AMS components. On the one hand, their modeling and design becomes extremely complex. On the other hand, their interplay with the rest of the system–on–chip challenges design, verification and test. The new technology trends bring up enormous challenges and opportunities for AMS design automation. This is reflected by an increase in research activity on AMS CAD worldwide. The purpose of this workshop is to report recent advances on AMS CAD and motivate new research topics and directions in this area.<br>The workshop puts the focus on the following topics:<br>1. Design verification and test<br>The embedding of AMS components in system–on–chips challenges the verification of overall system performance with the clash of two completely different worlds. New methods for verification and test of AMS components shall be presented and discussed.<br>2. Statistical design<br>Ever increasing variability in manufacturing requires efficient methods for design and verification of high–sigma designs. Where are we and were should be go? The workshop will discuss recent trends from application point of view.<br>3. Constraints and layout design<br>AMS layout is one of the last frontiers where design automation has not yet found its way into industrial application: industrial analog layout is still mostly “polygon pushingâ€Â. But recently, strong research effort has been spent to improve analog layout automation. We will review and discuss new layout approaches.<br>4. Analog benchmarks reloaded<br>At ISCAS`84, several researchers made an attempt to provide benchmarks for analog design. 30 years after, analog design automation still lacks reproducibility of algorithmic results by the community. We want to discuss pro and con of analog benchmarks and consider a potential initiative to create a benchmark suite for analog design problems.<br>
City
San Jose
Country
United States
Deadline Paper
Start Date
End Date
Abstract