EIPC Technical Snapshot: A European Roundup

Estimated reading time: 16 minutes

Opto-electronics on PCBs
The final presentation of the morning session came from Richard Pitwon, CEO of Resolute Photonics in the UK. As a technologist with many years of transformational expertise in optical and photonic system interconnect and integration for data-communications applications, he gave an energetic and informative introduction to the increasing role of opto-electronics on printed circuit boards, with particular regard to technologies we could expect to see in the next 5-10 years as optical communication migrated onto the PCB and right up to the chip.

He began by discussing the driving trends in hyperscale data centres, enormous environments comprising at least 100,000 servers and their associated technical infrastructure, required to support the mass scale requirements of data and cloud computing. Hyperscale has become the dominant form of data centre in the world; examples of the largest are those operated by Amazon, Microsoft, and Google. Of the many communication protocols employed, Pitwon has chosen the trend in Ethernet speeds to illustrate the exponential upsurge in the consumption of digital information causing a rapid evolution in data rates. And the proportion of machine-to-machine communication within the data centre greatly exceeds that of machine-to-user communication.

The evolution of data rates is driving the migration of optical interconnect into data-centre systems to overcome bottlenecks resulting both from the increase in data rates and the lengths of conventional internal interconnect which could be several hundred metres. Moreover, separating servers into their constituent compute and memory resources by optical disaggregation allowd them to be allocated as required by the needs of each workload, and distance agnostic links enabled high bandwidth communication bridges so that resources can be dynamically assigned and traffic and compute loads can be balanced across the data centre.

A further consequence of increased data growth is the need for an increase in switch bandwidth. Pitwon explained the concept of advanced “spine-leaf” optical switching architecture and its benefits compared with the conventional “fat tree” system.

The migration of optical interconnect is progressing from external optical connectors on the front panel to system-level integration, from front-pluggable transceivers and mid-board optical modules to chip-level optical interconnect with co-packaged optics as the ultimate objective.

He defined three categories of optical circuit board: fibre-optic flex-planes, polymer waveguides and planar glass waveguides, and showed examples of each in proprietary systems. The results of a major North American round-robin project on polymer waveguides are shortly to be published, and Pitwon referred to a series of European projects in which he has participated.

He concluded his presentation with a glimpse into future developments in quantum computing and the characteristics of its ecosystem. He predicted that future hyperscale data centres will increasingly incorporate quantum computer and communication nodes to complement their capabilities.

PCB Market Update
The afternoon session, introduced and moderated by Martyn Gaudion, managing director at Polar Instruments in the UK, began with a PCB market update and outlook from Dr. Shiuh-Kao Chiang, managing partner at Prismark Partners in the U.S.

“We just passed through an amazing year,” Chiang said. Prismark has almost finished its analysis of 2021 data for the electronics and PCB industry. Many companies have not yet fully released their earnings, but the overall picture of 2021 is becoming clearer. “Even though we’re still in the later stages of the pandemic, the amazing thing last year was the overall growth of the electronics supply chain, and the momentum that is happening across the board,” he said.

Chiang listed the highlights of the 2021 PCB market, which achieved 23.4% growth and exceeded $80 billion globally, due to recovered demand, product upgrades, inventory restocking, average selling price increase, and a weak U.S. dollar. The packaging substrate market was very strong, especially for flip-chip-ball-grid-array, flip-chip-chip-scale-package, and antenna-in-package and system-in-package components.

Notebook personal computer demand remained strong in the first half of 2021, aided by changes in server platforms. Automotive electronics recovered, fueled by the transition to electric vehicles and advanced driver-assistance systems. Demand remained strong for industrial electronics, medical wearables, gaming, smart consumer electronics, and internet of things.

A characteristic of the 2021 PCB market was surging revenue and profit growth in packaging substrates, although strong demand, high material costs, fractured supply logistics, increasing inventory, and tightening cash flow squeezed the margins of most PCB companies. Other important issues influencing the 2021 market were supply availability and capacity constraints for PCB raw materials, currency exchange rates, long lead times for critical processing tools and supply shortages for selective semiconductors and other devices.

Chiang expects more caution as the industry enters 2022 against strong headwinds—inflation, high interest rates, geographic conflicts, increasing materials costs, supply chain logistics, labour shortages, and other issues. But even in the later stages of the pandemic, the recovery momentum would continue and the electronics market would continue to grow, although at a more moderate rate, while substrates would continue to register double-digit growth.

To support the strong growth in the PCB industry through 2021, the upstream materials industry has suffered extreme turbulence: supply issues, cost issues, selling price issues. Prismark observed that relatively low-end commodity-type FR-4 laminates have increased in price by at least 50%, reflecting price increases in copper foil, glass fabric and resin; several suppliers in China and Taiwan have achieved revenue growth of over 100%.

Two factors that have affected the whole industry during the pandemic are a weak U.S. dollar and a change in purchasing mentality from “just-in-time” to “just-in-case” resulting in double-booking as people sought to secure supplies.

The semiconductor industry is in the middle of a five-year “super-investment” cycle, 2019 to 2024, with an expected spend of over $20 billion to build capacity for substrates, particularly those based on Ajinomoto build-up film (ABF). This increased capacity will create additional demand for associated assembly, and test and inspection equipment.

The market for HDI PCBs is expected to grow in 2022 as more non-consumer applications such as automotive, high-performance computers, high-speed networking, and satellite communications require HDI products. The smartphone market is expected to grow moderately. The demand for computer and communications infrastructure would continue in the economic recovery phase, supported by fiscal policies and infrastructure-rebuild programs, and this would strengthen the market for multilayer boards. Automotive production would be re-established when component shortage issues were resolved.

Chiang commented that the value growth in 2021 for the multilayer sector has been partly supported by increased selling prices resulting from material price increases, but when the cost structure for basic double-sided, four-layer and six-layer is examined, it is clear that direct materials costs account for almost 60% of selling price.

Growth is expected in the market for flexible circuits, particularly on low-loss substrates for millimetre-wave applications. Additional applications in areas such as automotive battery management, displays, wearables and internet-of-things will also increase demand.

Although Europe has seen a growth rebound in 2021, China remains the principal manufacturing location for rigid and flexible boards and is attracting significant investment, particularly in substrate capacity. He remarked that Japan, Korea, and Taiwan also have strong growth in substrate manufacture.

Chiang closed his presentation with a note of caution. “We are looking at this year with tremendous uncertainty,” he said. “Overall, we expect that 2022 will still achieve mid-single-digit growth, and hopefully the situation will gradually improve after the first half.”

Final Finishing Technologies
The focus of the session moved from market forecasting to chemistry, as Dr. Britta Schafsteller, Atotech's Global Product Manager Selective Finishing, discussed future opportunities for final finishing technologies.

Her selective finishes roadmap charted established and next-generation processes, together with an R&D outlook for rigid PCBs, SLPs, flex, and flex-rigid circuits as well as IC substrates. She listed the key finishing challenges to be overcome for each category.

For circuits operating at higher frequencies, these include tight impedance control, lower trace roughness and nickel-free surface finish for improved insertion loss. For line and space downscaling, the requirements are for a finely structured surface finish, advanced process control and particle reduction for higher yield.

Additionally, flex substrates with reduced lines and spaces required bendable layers, improved layer homogeneity and cyanide-free immersion gold.

As input-output density of IC substrates increases, there are requirements for bump pitch reduction and improved coplanarity for bump arrays, improved compatibility with materials, reduction of the consumption of precious metals, and extended process lifetimes.

Schaftsteller described in detail the characteristics and recommended applications for a whole series of Atotech finishes: a universal process for electroless-palladium-autocatalytic gold, electroless-nickel-immersion-gold, electroless-nickel-electroless-palladium-immersion-gold, a new palladium bath for pure deposits, a mixed-reaction gold, an immersion tin for flexible applications, an electroless-nickel-immersion-gold and a cyanide-free gold for flexible applications, and an immersion tin for IC substrates.

A newly developed fully-autocatalytic tin with no limitations in plating thickness or substrate will open new opportunities, with potential applications in IC substrates and micro-LEDs as an alternative to solder printing as finer feature sizes make such printing impracticable.

R&D in PCB
The final presentation came from Thomas Cramer, project manager for research and development at ILFA in Germany. He discussed the significance of research projects for the PCB manufacturer.

Introducing ILFA as a well-positioned and successfully owner-managed PCB manufacturer, EN9100 accredited as an aerospace supplier, with its own R&D department and owning several active patents, he commented that ILFA re-invested 10% of annual revenue to achieve its roadmap goals. ILFA was not identified solely as a PCB fabricator; the company is very customer-orientated and keen to cooperate as technical consultant in projects from the concept stage, right through planning, CAD layout, PCB production, and assembly of components until the final product.

Current technology demands from ILFA’s market are for miniaturisation, signal integrity, thermal management in high-current and ultra-thin multilayer designs, and withstanding harsh environments. Inevitably these demands must be met under time and cost pressures. Sometimes it is sufficient to optimise existing processes and materials, but often it is necessary to invent or implement new technologies.

ILFA has set up a technology roadmap to set timelines on their progress. One project is solely focused on a glass PCB with combined electrical and photonic vias, with glass layers and electrical layers in 10-layer construction, with blind via aspect ratios up to 1:3.

Cramer discussed routes to achieving research and development goals. Obviously the first is to recognise a market need, and possibly work together with a customer or supplier. Could the objective be realised by optimisation of existing technology? “Most important is to concentrate on your strong points, don’t focus on your weaknesses, but advance and go forward,” Cramer said.

The first option is to tackle the project alone, internally, on your own timeline, at your own pace, and keep all the success to yourself. Many companies operating in ILFA’s field are doing this but it could be a smart decision to widen the circle, possibly with an institute that has expertise in that area. The really smart solution is to join a government project that will provide both expertise and funding.

The extent of funding varies with the size of the operation, smaller companies benefiting most, and there is the opportunity to work with several different partners. If the project is ultimately not successful the costs are shared, and with the government contribution, the costs to individual consortium members are effectively reduced. The strong points are the larger audience and the benefit of being included in the partners’ marketing strategies.

Cramer quoted some statistics that illustrate the enormous growth in China’s share of the world market over the previous decade, whereas the German industry has retained a small but reasonably steady share. He stressed the importance of “going forward together” and trying to strengthen each other so that this share can be maintained.

Gaudion moderated the question-and-answer session, and then thanked all participants, especially the EIPC staff for the smooth organisation of a memorable event.

• < Previous Page