The electronics industry has always been a competitive one, and one of the fastest changing too. The never-ending demand for tinier, thinner machines with advanced functionalities need advanced PCB and parts. However, companies mostly outsource this process, which results in longer wait time. But integration of 3D printing with the electronics manufacturing processes is acting like a silver lining for the industry, empowering the industry to create in-house prototypes of circuits and circuit boards. This reduces the expense, time and concerns about IP violations too.

This deep integration of 3D printing in electronics industry is pumping up the market. The global 3D printed electronics market size is projected to reach $1015 million by 2027 from $204.1 million in 2020 growing at a CAGR of 25.2% during 2021-2027, reports Market Study Report LLC.

I spoke to Mahesh Deshpande, Senior Director, Global Business Consultant, High Tech Industry at Dassault Systèmes. A digital transformation evangelist, Mahesh is a seasoned technology product solution development professional who is driven by customer value, business impact and vision. Mahesh holds over 24 years of experience in R&D, business development, product innovation consulting. We discuss about how 3D printing and 3D technologies are changing the electronic industry and providing it with more power.

How 3D printing assists in semiconductor manufacturing and deal with the semiconductor shortage the world is experiencing?

3D printing in Electronics and Semiconductor industry in particular hasn’t lived yet to the hype of mass production and the underlying complexity primarily because semiconductors is a mass volume game and additive manufacturing technologies are still cost prohibitive and time consuming for volumes. A large semiconductor company can supply anywhere upwards of 200 million plus products everyday worldwide, and additive technology is more suitable for low volume, more complex things.

That said, there is a much larger potential for application of 3D printing into semiconductor fabrication equipment producing these volumes. These are extremely complex equipment, and are fully automated. They are multidisciplinary and the underlying complexity of the parts and the low volume really lends itself to 3D printing. There have been several efforts underway to bring the additive, particularly the metal additive technologies, to create highly specialized parts, say for the photolithography systems. As the shortage gets more acute, the need to run these fabs continuously with extremely high reliability is very important.

Any need to very rapidly diagnose the underlying issue in a semiconductor fab equipment, make a replacement of a defective part and bring it quickly back to the fab is a perfect use case of 3D printing. You can have a very fast turnaround time for the replacement, maintenance and repair services.

To fill in the existing gaps, we need to manufacture semiconductors faster. However, we cannot compromise with the testing and quality checks. How is 3D technology ready to help deal with this challenge?

The current shortage is strictly not a product design problem but a manufacturing and supply problem. The delivery lead times in semiconductors are inherently long due to the complexity of manufacturing process. A typical large company has to deal with, say 500 types of packages, 400 upwards of different types of chip ice, multiple technologies and 2000 plus process steps and the lead times can be several months long (20 to 26 weeks), of which a much larger portion is spent in the front-end process, up to 20 weeks and over six weeks in the back-end for assembly and test.

To deal with such complexities, a large amount of transformation is still needed. The superficial problem of demand patterns is still forthcoming and they’ve not being able to cope with it. But in this manufacturing and supply chain of semiconductors lie tremendous opportunities to have the right visibility and control. For example, from a manufacturing flexibility standpoint, you have to deal with the allocation of the various semiconductor chips to the various plants and the processes have to go through a large number of steps. All of these can be modeled in a model-based engineering paradigm, which gives a complete view of the bill of materials as it evolves through the various manufacturing steps at the front-end and the back-end.

Then there is the issue of multiple sourcing from various suppliers in order to mitigate the risks and achieve the right volumes to model that in your product definition. There is also the need for cross manufacturing – being able to produce the same chip in more than one fab, being able to qualify that upfront, technology- and customer-wise, routing of the product through the supply network dynamically depending on whether you want to prioritize the supply order for a very strategic customer or prioritize the routing through the network of a very high margin product.

A Model-based engineering and model-based manufacturing approach can provide the entire flexibility, the associativity all the way from design to manufacturing to supply chain. We can have a model-based supply planning approach which can help better connect the demand from the various customers and partners with the available supply networks. It also enables producing the right volumes in the right fabs according to the right KPIs.

In essence, there is tremendous opportunity to improve the responsiveness of the semiconductor fabs and the OEMs to address this prevailing shortage. However, redesigning the products is also necessary, taking into account alternate suppliers to components that are facing acute shortage currently. Within this model-based engineering approach and applying various simulation technologies, the companies can very quickly re-engineer or incrementally engineer the product and validate it and bring it faster to the market.

With the tremendous adoption and growth of connected devices across industries, what opportunities lie for 3D technologies in IC designing? What are some of the challenges and trends you foresee in this space?

Connected devices will be more pervasive in terms of new form factors, the range of wireless connectivity, the sensing abilities it will have, and the underlying need for a sustainable power supply to make them more autonomous. This requires the next-level IC design, and needs to go beyond the current nanoscale limitations with newer technologies such as 3D packaging with the System-in-Package with 3D stacking of dice on a smaller footprint, or Substrate like PCBs. One brilliant instance is the iPhone X where Apple was able to reduce 30% volume of the motherboard while all the chips remained as usual.

Such things require an integrated system level design approach. We must consider a holistic design from the chip to the package to the multiple PCB boards, integrate other important designs like battery design from a power system standpoint that brings multi-physics phenomena. Hence testing becomes integral including from a structural integrity, thermal behavior, electromagnetics standpoint.

One other interesting shift currently underway is the gradual shift of semiconductor IP and design capacity in-house, unlike totally relying on the external semiconductor providers. This move is because of the need of specialized chips required for everything, like chips for data centres, AI, ML, automotive, battery power management, networking and others. So the underlying challenge is the need to build the competency to develop chips in-house for differentiation, to have a strong roadmap for your products and time to market. Just as every company became a software company, virtually every manufacturing company will become a semiconductor company from a design standpoint in the next decade. This requires new workforce having strong electronics design competencies and system design competencies in-house. It also requires new methods of development of electronics much rapidly unlike the current ones that have long lead times in the semiconductors and the micro-electronics. On the other hand, the use of IP in building semiconductor grows seven times in every 10 years, and so this IP management and the modularity is going to be critical. These are some of the trends I foresee in the electronics and semiconductor design as the connected devices get pervasive.

Indeed with taking semiconductor manufacturing in-house, there will be a plethora of challenges these companies will face. Will 3D technologies help them save costs, while become faster in reaching the market?

Currently the use of semiconductors is driven by a mix of specialization and differentiation, relying on some key semiconductor design partners, chip providers and bringing some in-house. But this is also going to be a part of an IP marketplace just like it happened in the mechanical space. The ability to very quickly find the right supplier for a specific type of mechanical or electrical component, an ecosystem for IP for semiconductor and electronics design is going to grow. With that comes the need to have a better visibility on what kind of IP is coming into my product, how mature it is, who are the suppliers, compliance, royalty and licensing. All of these have to be traced from the incoming sources of the IP all the way into your product lines and the life of the products.

There is also a strong need for validation of IP from a performance, reliability and manufacturing perspectives. This is not just a transactional issue any longer but is the core of the innovation and the engine to give you faster time-to-market. We need different solutions along the way to manage the integrity of the IP, the modularity of the products based on the IPs, the continuous verification and validation across different kinds of IPs, and then the handover of the design for the testing and the new product introduction cycles. This could be a digital innovation platform that can lend itself to the semiconductor based innovation.

Electronics skin is almost a reality. What role has 3D technologies in achieving thinner chip, and if there are any challenges?

Semiconductors are not only getting thinner but also as a part of the whole product package, they are getting a lot more flexible and becoming higher density products. This brings the need for ultra-low bar processors, communication, chips, memory, all of them integrated but still mounted on a printed circuit board to form a system level solution. It also needs to have the right interfaces, the battery, sensors and more. When you put all of these together, testing becomes all the more difficult because of the complexities and the high functional integration.

In terms of use cases, we can think of the need to design new materials, whether organic or inorganic, silicon or other substrates. Also the antenna remains a key component of the e-skin or other connected products whether the antenna is a chip level one or a PCB mounted, or the antenna could be custom design. Hence we need to have an optimum placement of the antenna from a reception performance standpoint. We also need to assess all kinds of electro-magnetic interference and compatibilities through an electro-magnetic simulation.

The antenna interacts a lot with the rest of the device, and doesn’t act in isolation. It  interacts with the structure of the device creating a resonance or also with the battery. All of these have to be assessed through a holistic multi-physics electronics simulation. There is also the human aspect to consider in the e-skin and wearables application. So we have to perform a SAR (Standard Absorption Rate) kind of simulation to check the impact of the antenaa radiation on the human tissues. Hence there is a tremendous need for various 3D-based virtual technologies to deal with these complexities and miniaturization, and to be able to test multitude of scenarios that are just not possible to test in a physical way. We have a front load them in the virtual world.

Let’s talk about the communication industry. With all the 5G revolution being work-in-progress, there still exists multitude of challenges. Is there a collaborative way in which 3D technologies can enable a better 5G experience and service to the customers?

5G will have a hybrid mode. We will see many converged network sites with a mix of 4G and 5G. A lot of efforts is being put for network densification depending on the topology and urban landscape, among others. There is also a configuration density; we will have macrocells to small cells to distributed antenna cells to serve various capacity and latency needs. There is also regulatory pressure for carrier aggregation and spectrum sharing. Hence maintaining quality of service and being able to recover quickly from the outages and planned downtime is really critical for customer satisfaction.

Hence it is important for the telecom operators to have a complete view of the network site, including the physical infrastructure, and logical layer (network layer). We can envision this virtual twin of the network sites to allow issue detection, such as degradation of service, or being able to plan new network upgrades and do all of these without excessive site visits and interruption to the ongoing service. Newer technologies like 3D help in having an as-built view of the network information and having realistic 3D work instructions for the service delivery, whether it is for upgrades or some quick maintenance.

This will empower the telecom operators to evolve from a document and physical monitoring based approach to more model-based and hybrid- virtual+physical -approach.

Our 5G network equipment players are helping the operators in the network rollout and the managed services. And in this issue of being able to quickly roll out a site, technologies like BIM can get a complete view of the various construction deliverables, all the design assets in such a way that the telecom company is able to perform these network upgrades and expansions with minimum site surveys.

Further, the active antennas and various configurations can bring extraordinary complexities to 5G. You cannot test these antennas just in the lab setting because the antenna is subject to the real world environmental constraints like wind loads, rain and site constraints depending on the neighborhood around the site. So the form factor of the antenna, the coverage, throughput, interference and more needs to be simulated in a dynamic environment. With 5G, some new data streaming technologies are being added, such as massive MIMO and beam forming behaviors. The good thing is that all of these can be tested without incurring the physical prototype testing or on-site testing with a holistic multi-physics simulation approach. Various use cases can be tested all in conjunction.

Anything that you would like to add in terms of 3D in the world of electronics?

This notion of Virtual Twin, and Model-based Enterprise, even though it began in the industrial manufacturing world in aerospace and automotive, has rapidly made its way into the world of high-tech and electronics. We now have the maturity of looking at electronics as a whole system and manage the evolution of that system, not only in the individual pieces of mechanical, electrical, electronics but also its overall performance. Being able to deliver this with a faster time to market, with the right quality and right cost remains a challenge. So the time to apply the 3D virtual technologies – Virtual Twin and Model-based Enterprise – is now and that the electronics companies – whether they are OEM or in the supply chain – can tremendously benefit from taking this platform based approach to the R&D of the new electronics-based products, but also continue that downstream in the value chain into testing, manufacturing and supply chain.