The automotive industry’s interest in semiconductor products is increasing daily, and expectations will only get higher in the coming years – as Yole Développement illustrates in its Trends in Automotive Packaging 2018 report. Driven by the adoption of more electronic components in vehicles, the semiconductor industry has begun facing very high demand. In this context, advanced packaging technology will become increasingly preponderant, and the automotive industry needs are indeed getting more specific regarding regulations, qualifications, and testing. Cars have become the new “consumer” products, and therefore it is essential to adapt current consumer technologies to automotive in order to drive innovation.
Although today the main platform in term of units is the wire-bond ball-grid array, which represents half of the market, advanced packaging platforms like flip-chip and fan-out are steadily finding their place. Autonomous driving could be the main future trend for pushing towards transfer 3D package, from established markets to the automotive market. Better computing performance and interconnection density are claimed by this application, but reliability is an automotive keystone which could be hamper it in the beginning.
Lauranne Chemisky, Market and Technology Analyst at Yole, recently interviewed Glenn G. Daves, Vice President of Package Innovation at NXP Semiconductors, to learn more about the automotive market and understand NXP’s vision and perspectives on the future of automotive advanced packaging.
(Source: Trends in Automotive Packaging 2018 – Yole Développement)
Lauranne Chemisky (LC): NXP is one of the major players in the automotive semiconductor industry. As one of the bigger IDMs, how would you describe the noticeable changes in the automotive market these past few years? What is your view regarding the impact of the “4 trends” identified by Yole: electrification, autonomy, connectivity, and comfort?
Glenn G. Daves (GD): With NXP’s position as the largest automotive semiconductor company, we tend to get a broader view of the sweeping trends that are affecting the industry. We generally put these into three categories: connected infotainment, enhanced ADAS (advanced driver assistance systems) moving towards autonomy, and electrification. In the longer-term, all three will bring major changes to cars, to the companies that produce and support them, and to our lives. We’re very excited about these changes and how NXP will be able to help propel them. One other interesting trend not typically mentioned is how industry relationships are changing. We are seeing auto OEMs take a more central role in developing the systems that go into their cars, rather than relying solely on the Tier 1 component suppliers. We are also seeing many new players enter this space, seeking to sell their products in this growth market.
LC: In your opinion, what trend (between electrification, autonomy, connectivity, or comfort) is the most important in today’s automotive industry? Everyone talks about autonomous vehicles, but in revenue the main trend seems to be electrification. What is your opinion on this evolution?
GD: All of these trends are important, but they are also clearly on different introduction schedules. Arguably, connected infotainment systems have led the way with the earliest adoption into mainstream vehicles. Even entry-level vehicles today have infotainment systems that are far beyond what many luxury vehicles had just 10 years ago. I don’t think you can see that same kind of dramatic change in other vehicle systems yet. Between electrification and enhanced ADAS, these seem to be on similar introduction schedules. Cars are getting more electric and much smarter at about the same pace. But I think the path to full autonomy is a longer one. I was recently in Bangkok in the middle of a traffic jam, and based on this experience I have no doubt it will take some time before a car can efficiently navigate that kind of environment in a fully autonomous mode. Thus, my view is that, while electrification and enhanced ADAS are starting at about the same pace of introduction, vehicle electrification will likely complete first.
LC: How much does the automotive business represent for NXP, and how important is packaging in your automotive business?
GD: NXP is the largest automotive semiconductor supplier, and this market represents about half of our total revenue.
Packaging is, of course, important in all markets. But in the automotive market, there are unique challenges that perhaps render packaging even more important. The automotive environment poses very challenging use-cases, with extreme lifespans, on/off cycles, ambient and operating temperature ranges, and media exposure (especially regarding certain sensing systems). On top of all this is the overriding concern for safety, which drives quality and reliability expectations that are quite lofty. Thus the package is instrumental to the success of our products in meeting all of these requirements.
LC: We see more and more packaging technologies from the consumer market being adopted in the automotive market. What do you think about this new strategy? Do you think it will continue, and could it help consumer packaging players enter the automotive industry?
GD: We do indeed see the migration of technologies into cars that were formerly limited to consumer or industrial spaces. In fact, NXP was among the first companies to introduce complex flip-chip, fan-out, and WLCSP packages into cars, and especially into automotive safety segments. This trend will certainly continue. However, I should be clear that consumer packaging technologies are, for the most part, completely unsuitable for the automotive space without significant changes made to the BOM, the assembly processes, and/or the manufacturing controls of the factories that assemble them. These are typically dramatically different from the original starting point. Because of this, I don’t think consumer semiconductor companies have a “leg up” in some of the newer, advanced automotive spaces. Rather, I think they have a steep learning curve to climb. Many of these companies understand the more challenging requirements of the automotive market and are fully engaged in climbing that learning curve. But many others do not, and thus are likely to experience some very unpleasant surprises.
(Courtesy of NXP)
LC: Yole sees more packaging assembly being outsourced to OSATs. What in your opinion drives this trend, and what type of packages are being outsourced from IDM to OSAT? With OSATs growing in the automotive segment and representing in 2017 almost 40% of total packaging revenue, how would you estimate the market-share evolution between IDMs and OSATs?
GD: The industry trend over many years has been “increasing market share at OSATs”. Much of this increase has been accomplished by OSATs purchasing IDM assembly factories. With rare exceptions, the full suite of industry packages are available at OSATs, and most IDMs outsource all package types. Turning to automotive, the field is more limited. Most OSAT sites have no or very limited experience with the automotive market. This is especially true when looking at more complex packages. Given the market trends, I would expect that OSATs’ overall market share will continue growing, including in automotive packages, though these will lag somewhat compared to other market spaces.
LC: Yole sees Amkor and ASE as the main OSAT players in automotive packaging. Do you think that, despite regulations and lofty quality expectations, smaller OSATs could enter the market in the future?
GD: Many OSATs are playing in the automotive space today. These include both large and small companies, and I expect that more will enter given the long-term growth prospects for automotive semiconductors. However, I do agree that the automotive market is a space where larger companies with more resources will have an advantage over smaller companies. Again, especially with more complex packages, the challenges are steep. To succeed, you typically need larger R&D teams with deep expertise in relatively narrow technical sub-specialties. But size is not everything. Probably even more important-and this is an area where a company of any size has the ability to outperform its competitors -is the need for absolute organizational and leadership commitment to succeeding. Every company in the automotive space will face really steep challenges: for example, having to solve a 1-ppm quality problem. Only those leaders that are truly committed will be able to drive themselves and their teams to make the effort necessary to address an issue like this. If your first thought is, “Oh, it is only 1 ppm”, you and your team will never get there.
LC: LiDAR is one of the new devices that will enable more autonomy in driving, Yole sees LiDAR achieving the highest growth in unit and packaging revenue. What challenges does LiDAR face?
GD: LiDAR systems are very low-volume today, but they are expected to grow in the future as the technology matures and transforms into solid-state. Radar systems, on the other hand, are very high-volume today, are moving into their 3rd generation, and are increasingly being installed in mass-market cars in applications such as adaptive cruise control and emergency braking. NXP produces automotive radar systems, and we’ve been active in these systems from the very beginning. The packaging challenges are high: signal integrity at 77 – 79 GHz and transforming advanced packages (and the factories that produce them) into something suitable for automotive safety applications, along with the “normal” packaging concerns of power integrity, thermal performance, and mechanical performance. But by overcoming these challenges, NXP has had the privilege of enabling transformative applications that revolutionize driving and save lives. It is truly a rewarding endeavor.
LC: Today, WBGA is the main package type used in automotive, but flip-chip is growing and embedded technologies are considered as disruptive solutions in power. What are your expectations for these new packages in automotive?
GD: Actually, lead frame packages continue to dominate in the automotive space in terms of volume. I expect this dominance to persist in the future, as they have many advantages for many applications. But, you are correct that, at the higher end, where wire bond BGA packages once dominated, you now find flip-chip. We are also seeing fan-out and WLCSP packages in cars. These advanced packages are here to stay. The reason they are used is because the bus speed, signal integrity, power integrity, total i/o, and heat transfer demands are rising dramatically both in infotainment and in ADAS/autonomous driving systems. In the power semiconductors space, certain embedded package types, with thick Cu conductors, offer performance advantages. Of course, all of these advanced packages come with a cost.
NXP is getting much better at asking and answering the question of “how much performance is good enough?” Especially when “better” may cost twice as much, “good enough” is a very attractive option. We use advanced packages when we need to, but we also think that wire bond BGA, right down through lead frame, will be around for a long time. And we are actively developing these lower-end packages to make them more capable and thus extend their lifecycles.
LC: The flip-chip platform seems to be getting more popular in many automotive systems. Do you also see this technology being increasingly adopted and achieving higher-grade qualification so that it can be used in any car application?
GD: Yes. For NXP’s products, flip-chip is expanding beyond infotainment systems and into safety applications like vision and radar controllers. We have announced products in flip-chip packages that will meet AEC Grade 1 requirements. Others are following down this path too, so over time the industry will expand the role of flip-chip packaging in cars.
LC: The use of organic substrates seems to be growing. What are the main technical limitations for this type of package?
GD: NXP uses organic substrates extensively in its packages, and has for many years. Of course, there are many different kinds of organic substrates, from low to high-end, and each has its own limitations. But, looking at the category in general, we do not see many technical limitations for our product spaces. For example, we have successfully qualified substrate packages for the most stringent automotive environments (AEC Grade 0). Two current areas where we are focusing to extend substrates are, first, millimeter-wave applications like 5G and radar. Here, we think adjustments can be made that will improve signal integrity at these high speeds. The second area is manufacturing controls. As the march for improved quality continues, we are working with our substrate suppliers to adopt best practices for SPC, cleanliness, etc. which will drive more consistent, higher-quality output. We don’t see either of these as an insurmountable challenge and we are confident in the ability of organic substrates to serve us well into the future.
LC: How is NXP developing new products or solutions to compete in the automotive industry at the packaging level? In the automotive IDM landscape, what differentiates you from your competitors? How do you envision NXP’s future in automotive?
GD: NXP is continually developing new products across the full spectrum of automotive applications: from tire pressure monitors, to the electronics for braking systems, to a full portfolio of infotainment and radio products. In each of these systems, NXP’s packaging is an integral part. We are continually pushing the envelope in squeezing more capability from traditional and advanced packages, with higher quality and reliability. We differentiate from our competitors product by product, but also in that we bring a comprehensive and integrated suite of products to the automotive space, and this includes packaging. NXP is extremely committed to the automotive market. We were the world’s first supplier of automotive electronics, and we plan to continue to grow our contributions in this space.
Glenn G. Daves is Vice President of Package Innovation at NXP Semiconductors. He is responsible for package design, package technology development, and assembly process development in support of NXP’s full product portfolio. Prior to its acquisition by NXP, Glenn led packaging and printed circuit board development for Freescale Semiconductor. Prior to that, he led global packaging product and technology development at the IBM Corporation. He has also held leadership positions in project management, test and burn-in engineering, and assembly manufacturing engineering. Glenn holds twenty-eight U.S. patents and has degrees from Brown University, the University of Illinois at Urbana-Champaign, and Alliance Theological Seminary. He serves on the board of trustees of Nyack College and on the National Leadership Council of World Vision U.S.
Lauranne Chemisky is a technology and market analyst in the Semiconductor & Software Team at Yole Développement (Yole). Lauranne is currently engaged in the development of market research reports as well as customized services for clients. She is able to leverage her technology training and experience in the fields of materials and semiconductor manufacturing processes for advanced packaging applications. Previously, Lauranne worked at Apple in the Softgoods Product Design Team as a material development engineer (CA, USA).
Lauranne holds MSc in Materials and Polymers Science (ITECH Lyon, FR) and Master in Management and Marketing for new technologies and innovation (EM Lyon Business School, FR).
Trends in automotive 2018
OSATs are gaining benefits from advanced packaging in automotive.