Stringent High-speed Requirements Pose Technology Challenges

Estimated reading time: 2 minutes

The ever-increasing demands on printed circuit boards to satisfy the needs of tomorrow’s products means that PCB manufacturers must continuously evolve and react to a wide variety of technological and market requirements such as: Functional density: Finer features, higher density, and increased layers; signal integrity: Higher frequencies, driven by higher data transfer speeds and increased data needs; material properties refinement: Must meet performance as well as environmental demands; Smart manufacturing: Automation, robotics, artificial intelligence, augmented and virtual reality, and machine learning are all part of the future PCB manufacturing floor; and time-to-market: Increased urgency to facilitate new design implementation, product introduction and qualification, quicker market differentiation, and survival.

To meet these stringent requirements, driven by expanding data needs and increasing transfer speeds, the electronics industry turns to high-speed PCBs and substrates. This article, based on the iNEMI Roadmap1, looks at some of the future requirements facing high-speed PCBs and substrates and discusses technical needs, gaps and solutions.

High-speed PCBs

The roadmap’s examination of high-speed PCBs focuses on solutions for high-speed digital applications such as high-performance computing in data centers. In this context, rapid edge rates, as encountered in signal rise time or fall time (whichever is less), define the signal as “high speed.” A rise or fall time of 1 ns implies a total signal bandwidth of 1 GHz or more, even with much lower switching speeds.

Technology evolution is driven by requirements to be:

• Faster: Interconnect speed will often be the limiting factor, particularly at the package level. Per-lane signaling speeds need to continually increase and that implies extra signal bandwidths. Issues such as losses and signal integrity will be critical.
• Smaller: Increased miniaturization in the core compute processing units leads to finer features and increased functionality and/or capacity on a given PCB, which leads to higher layer counts, as well as a reduction in all geometries.
• Hotter: Even with improvements in energy efficiency, heavier computing workloads from paradigms like machine learning and increased overall capacity will drive up power consumption. PCBs will have to accommodate higher power delivery and new approaches to thermal management, such as immersion cooling for data centers.
• Greener: PCB technology must become more sustainable, with the development of more energy-efficient manufacturing processes and with new materials and stackups supporting increased recycling and reuse.
• Cheaper: For infrastructure electronics, “cost” must include more than the bill of materials; it must focus on the full lifecycle. Improved reliability is a major objective to drive down lifetime costs.

To read the rest of this article, which appeared in the September 2024 issue of PCB007 Magazine, click here.