Better Chips, Better Cars

The foundry perspective on automotive chip manufacturing.


There are literally thousands of electronic components in a new car, and those numbers are only going to increase as cars become smarter, safer, greener, and increasingly connected.

As automakers and Tier 1 and Tier 2 companies shift their focus from mechanical to a combination of mechanical and electrical, there is an ongoing race among fabless companies to come up with innovative technologies for everything from ADAS to infotainment to better connectivity.

According to Gartner, automotive semiconductor revenue will grow at about twice the rate of the overall semiconductor market through 2020. From a foundry perspective, having the manufacturing capabilities and technologies to help customers achieve reliability and meet stringent automotive quality requirements is key.

Stringent standards
As with any safety-critical device, the goal is zero defects for devices. These devices need to operate flawlessly for years under some of the harshest conditions imaginable. A Grade 0 device needs to operate effectively between -40°C to +150°C, while a Grade 1 device needs to be able to operate within -40°C to +125°C.

While these types of standards have been in use in aerospace and military applications for some time, it is new to many fabless semiconductor companies who want to participate in the fast growing automotive market. Suppliers of automotive IC components not only to have to meet automotive qualification, such as AEC-Q100, but also have to meet TS-16949. Many vendors add BCM certification as part of their supplier vetting requirements.

TS-16949 quality management system specification is intended to be applied throughout the entire supply chain to help drive toward zero defects and reduction of variation. The updated IATF 16949:2016 has a sub-clause that now requires the organization to explicitly meet customer requirements beyond AEC-Q100.

Process nodes will vary
Moreover, that testing will have to address a wide range of components using different processes, many of which are still evolving. There has been much discussion about the optimal process nodes for automotive, but the reality is that no single node will suit every purpose.

Most people are familiar with automotive buzzwords like V2X, infotainment, artificial intelligence, ADAS, and body control. When you dive deeper into the IC level, there are hundreds of components from a variety of nodes, such as the MCU, touch controller, display driver, wireless module, application processor, battery management IC, image sensors, etc. Some are driven by cost, while others are driven by performance or technology limitation. Massive integration of all of these features into a single silicon is not going to happen. On the contrary, it would be to the benefit of a semiconductor vendor to drive as much silicon content into the car as possible.

When it comes to node selection, there is always a tradeoff between performance and cost. Not all foundries are created equal. Each foundry establishes its own strength and its unique position in the market. Some are focused on building bleeding edge technology, some are focused on building III-V, and some are focused on memory.

For example, UMC is focused on building a full spectrum of technology and is known for being flexible and step beyond standard offerings. Not all foundries put in the strategic investment like UMC does to make all fabs to comply with rigorous automotive quality standards.

As ASPs for bleeding edge technology continue to follow the declining curve, the development costs of these bleeding edge products have gone the complete opposite way. A 14nm tape out could cost millions. This financial barrier is pushing a lot of niche players out of the market. However, 40nm and 28nm processes are very mature, therefore the cost is very reasonable. They will continue to be an attractive option for many applications, from both a financial and performance perspective.

Foundry of choice
Finding a foundry partner with automotive experience is a necessity in bringing products to market quickly. Your foundry partner should continuously add automotive qualified processes into the mix, and create a comprehensive and flexible automotive service package.

As bleeding edge technology gets more and more cost prohibitive for Tier 2 semiconductor companies, mature process technologies can offer exceptional performance and cost, and there is an abundance of IP available from the ecosystem. N-1 or N-2 nodes, like 40nm and 28nm, can fit a variety of applications.

As you consider your product roadmap and migration path, look for a foundry partner that has a deep breadth of nodes and specialty technology offerings to give you the flexibility to improve and be competitive. Achieving zero defects requires robust manufacturing processes, procedures, as well as training built into to the manufacturing environment.

UMC Auto is a comprehensive platform that consists of a broad portfolio of automotive AEC-Q100 qualified technology solutions, backed by robust manufacturing processes that comply with rigorous ISO TS-16949 automotive quality standards for all UMC fabs. In addition, UMC is selectively developing certified design models, IP, and Foundry Design Kits specific to the UMC Auto platform in order to fulfill the increasing pace of change in the auto industry supply chain, helping chip designers capture new market opportunities as the Internet of Things and increased use of sensors permeate auto applications.

Leave a Reply

(Note: This name will be displayed publicly)