contents

hardware
 
AgO Inc Launches AnXplorer

AgO Inc launches AnXplorer a new tool for optimizing analog and RF circuits. Starting with an unsized SPICE netlist, variables for device dimensions and a set of design objectives and constraints, AnXplorer optimizes device sizes by rigorously exploring the design space. By centering the design across all specified, process, temperature and voltage corners, AnXplorer achieves a robust design thereby enhancing the yield and improving the probability of first time silicon success.

The analog and mixed signal segment is growing much more rapidly than the rest of the semiconductor market. This trend is unlikely to change with the current massive growth in mobile data communications requiring both RF and mixed-signal design. By 2014, Cisco predicts that mobile data traffic will increase a staggering 39 fold from 2009 to 3.6 exabytes per month globally (source: Global Mobile Data Traffic Forecast Update 2009-2014). Despite this, analog design automation has remained largely unchanged for many years. Circuits are usually entered by schematic capture and device sizes are manually tuned over many transistor-level simulation runs.

AnXplorer can fit into commonly used design toolsets as it works with industry standard SPICE netlists and simulators. A designer can create an unsized circuit, design variables and objectives. AnXplorer will then generate an optimized, centered circuit that meets or exceeds the design objectives across all corners specified by the user. It uses a new multi-algorithmic optimization strategy aided by an expert system.

A major challenge facing developers of analog and mixed signal ICs is the frequent failure of analog circuits to meet their requirements. Design respins not only significantly inflate the IC's development costs but also adversely impact the final product's market entry.

Users can control the way optimization takes place by defining a hierarchy of design objectives. Unlike weighted-based handling of multiple objectives, the tool tries to meet high priority objectives before optimizing lower priority ones. AnXplorer's design exploration database records all points during the optimization process. This enables designers to undertake "what-if" analysis or to trade off conflicting objectives if all objectives cannot realistically be met.



write your comments about the article :: 2010 Computing News :: home page