Mixed-signal applications are among the fastest growing segments in the electronics and semiconductor industry. Applications in mobile communication, networking, power management, automotive, medical, imaging, safety and security require a very high integration of analog and digital functionality at system, SoC and IP levels.
Unfortunately, compared with the advancement of digital designs over the past decade, the state of art analog design is significantly lagging behind. For example, the throughput of microprocessors doubles every 1.5 years while it takes three times longer to achieve the same advancement for analog designs. Another big roadblock for analog designs is the power consumption. According to Boris Murmann, professor at Stanford University, the equivalent digital gate count in terms of power consumption for a 10-bit ADC at 0.13 um is about 100K, and this number grows almost exponentially for larger ADC and modern advanced nodes.
A new circuit design technique, digitally assisted analog (DAA), delivers a promising solution to address the performance and power challenges to further expand the scope of analog designs to meet today's application requirements. Let's use a simple ADC to explain the concept of DAA:
Figure 1 shows a conventional ADC and Figure 2 shows a DAA style ADC. In Figure 2 a conventional, high performance, power consuming ADC is replaced by a very simple, low-power ADC, followed by a digital post-processor to apply corrections to the output to achieve the same accuracy as the conventional ADC. Compared to the conventional ADC, the DAA ADC has a significant benefit in terms of power and area.
In addition, DAA style designs are easier to port to advanced nodes since majority of the computation task will be performed by the digital post-processor which typically demonstrates an even larger advantage in power, performance and area (PPA) at advanced nodes. With the increasingly wide usage of embedded processors, such as the ARM Cortex-M series, designers can achieve additional benefits in terms of productivity and flexibility thanks to the great software capability of such processors.
The above example just illustrates one specific approach for DAA circuits. In general, in DAA circuits, the assisting digital logic is used to monitor analog performance through the different stages of the operation and to adjust parameters of the analog circuits (such as bias, resistance, capacitance) through calibration loops to meet overall design objectives.
We have seen significant advancements of DAA designs in recent years from the design community, and its proliferation signifies a new era of mixed-signal design. By replacing more and more analog circuitry with digital counterparts to achieve the ever more aggressive PPA targets, we foresee an explosion of new mixed-signal design starts. As a result, the industry is demanding a true mixed-signal design methodology for design, verification and implementation to meet the requirements of such design styles. In the follow-up blogs, we will talk more about how the Cadence mixed-signal solution is best positioned to meet such new mixed-sign design challenges and how you can learn more by joining us at DAC.