There’s a growing consensus that implementing successful AI and ML at the design level doesn’t lend itself to a one-size-fits-all design solution. To remedy this, many developers create their own application-specific ICs (ASICs). These developers are claiming that various benefits can be obtained with these ASICs. These benefits include better performance, more operations per cycle, and a simpler and more deterministic design compared to a CPU or GPU. The article that follows provides a summary of these design techniques, as well as a link to the full technical paper itself.
In a recently published technical paper, Mentor engineer Neel Natekar writes about the importance of ensuring reliable design and verification of AI and ML processors for the purpose of delivering legitimate results:
Artificial intelligence (AI) and machine learning (ML) are seeing growing adoption in a wide range of applications, due primarily to the improvement in algorithms, advancements in hardware design, and the increase in data volume created by digitization of information. The integrated circuits (ICs) used in AI/ML applications are characterized by large parallel processing computation units, high power dissipation, and complex circuitry that can deliver maximum performance within a strict power budget.
Ensuring the reliable design and verification of these complex ICs is critical, since circuit failures in these chips can have major consequences for the validity of the technology and legitimacy of the results they provide.
