Increasing engineering efficiency will require lowering the AI expertise barriers for everyone in the chip industry.
Semiconductor manufacturing is going through massive transformational challenges driven by strong demand for advanced computing, fueled by AI, cloud, the electrification of the economy, and the need for compute power in data centers to support these applications. With the slowdown of Moore’s Law, more compute power will not be achieved by just increasing transistor density.
Not only is Moore’s Law slowing down, but the cost per transistor is not decreasing. In fact, it could be increasing for the most advanced nodes. While the benefits are harder to achieve and require more than geometry scaling, demand for these advanced nodes continues to grow.
The slowdown of Moore’s Law is also pushing semiconductor design in new directions including 3D, chiplets and complex hybrid packages. It’s not just a design challenge; it triggers major changes in the overall supply chain across the industry and requires new capabilities like traceability and complex analytics for test, forcing the entire semiconductor ecosystem to engage in a major transformation.
These challenges impact everything from design and manufacturing to the already complex supply chain business model. The supply chain for 3D hybrid packages will be transformed, impacting testing and the overall division of responsibilities across the entire semiconductor industry.
In recent years, the global supply of semiconductor devices attracted the attention of many countries that want to bring silicon production to their shores. The unprecedented level of manufacturing investment will help, though it will create additional capacity as well as inefficiencies. The path for profitable investments is uncharted as a major increase in fab efficiency will be needed for them to be competitive and likely require accelerated learning that only AI and digital transformation can deliver.
AI’s power to transform semiconductor design and manufacturing into the industry of the future cannot be overstated. It is already revolutionizing semiconductor design and manufacturing with tangible examples of companies deploying AI to make a difference in their business and bottom line. Companies are deploying a cloud-based AI data analytics platform to analyze large amounts of manufacturing, product or test engineering data. Without this type of automated solution, only a small proportion of these data sets could be utilized.
AI analytic platforms with sophisticated machine learning techniques to analyze large data sets are helping engineers be more productive and deliver better results. It’s a change from the traditional method where companies staffed their product design and test engineering groups using a budget based on a percentage of revenue. If a company had revenue in the billions, it hired more product and test engineering without knowing the engineers’ productivity. Without AI, they could only use simple reports and graphs to analyze the subset of data.
For AI to have an even greater transformational impact, it must be more pervasive than it is today. The dilemma is lowering the AI barriers for everyone in the chip industry, not just the few specialists. Among the many steps along the way is developing an infrastructure to reduce the barrier of expertise required to learn how to apply these techniques. Currently, few experts know how to use AI out of the thousands of workers employed by semiconductor companies around the globe. To do that, the barriers need to be lowered, making systems more integrated and streamlined. By doing that, the data will flow more seamlessly from one application to another, aligning semantic models of information to apply algorithms without the need for data wrangling experts.
The transformation will help make the entire global supply chain become more resilient by expanding sourcing options for critical products or test applications.
 |  |
 |  |
 |  |
 |  |
 |  |
