Moore's Law continues to hold dominance in the industry, and so we see much discussion around the opportunities presented by new geometries especially for Consumer Electronics (CE) and enterprise markets. After all, the volume of components demanded by CE and enterprise sectors is significantly higher, cycles rapidly, and is affected by often sudden demand shifts. The industrial and manufacturing side of the semiconductor and electronics industry holds growth potentials from new geometries and chip architectures as well, but these sectors face some different challenges. Bridging the two sides of the semiconductor and electronics industry is that both are forced to rely on fewer and fewer fabs due to ongoing chip manufacturing consolidations while navigating the maze of barriers to entry continue to rise, restricting new competition from offering alternatives.
Balancing requirements & drivers
One aspect of design engineering (DE) that we don't hear about as often, is where the drivers and resulting balance points are for industrial and manufacturing sectors. To gain some insight, I spoke with a senior project engineer at a leading industrial automation company for an update on what challenges DE is facing as next generation nodes push below 1x-nanometers (nm) and new architectures favor 3D-stacked chip solutions. As he notes, there are a couple of central requirements in DE that have become increasingly challenging due to die shrinks and stacking: (1) designs must first and foremost have a path to satisfy longer lifespans while meeting strict quality requirements; and (2) designs must be ruggedized for reliability to continuously withstand the harsh industrial environments.
Unlike CE and enterprise solutions, which generally max-out their lifespan at the five-plus year mark, whether due to component or user expectations, the same five-year point marks only the halfway point of the average 10-year lifespan that industrial components ought to provide. Typically, design engineers lifespan requirements will generally range from five to 20 years for sectors such as industrial automation, automotive, aerospace/defense, medical equipment, oil & gas (O&G), among others. These are core requirements and are nothing new to DE, he reminds us. The pressing question is, will the new architectures, while certainly designed to meet longer lifespans and tested ruggedized for harsh environments, actually meet up to the stringent requirements demanded?
The senior project engineer reiterated that the screening for component requirements, particularly for temperature and different environments, effectively sorts commercial from industrial grade. However, he continued, the new chip architectures and geometries, while holding great promise, are still under scrutiny when it comes to long-term reliability performance. Among the concerns he contends with are soft error rates and the different failure mechanisms with the new geometries, particularly over time.
From the perspective of DE, one issue is while newer architectures and stacking certainly offer many potential advantages, in shrinking these integrated circuits, the gate oxide gets thinner and therefore various failures can creep up. For example, one question is what the effects of voltage spikes might be and what damage might occur given that ever-smaller transistors may be weaker and the fact that some of the input protection may be removed in order to increase speed; another question is how to contend with ever-denser populations of transistors per square area resulting in increased leakage currents and power consumption which are huge concerns.
Pricing power shifts add to DE challenges
Beyond the reliability challenges for DE, given new component roadmap designs, engineers face design challenges in balancing functional versus pricing requirements for components. On the pricing side, unlike designing for commercial sectors, the equation includes higher-grade component and lower volumes, each driving up pricing. Commercial designers are able to leverage high volume demand to push chip manufacturers' pricing downward. When it comes to pricing, we all know that volume holds sway.
Among industrial and manufacturing sectors, there are important shifts changing the balance of power with chip manufacturers. Automotive manufacturers are gaining in their influence due to a significant increase in the number of electronic components. The increase in automotive electronics is driven by the increase in demand from consumers for on-board information and entertainment as well as from mandated safety monitoring and reporting. As a caveat, although automotive is now able to command better component pricing within the industrial sector, they still lag in negotiating power compared to commercial demand for the components destined for on-board infotainment (e.g., memory, MCUs, MPUs, etc.). So while some pricing is favoring automotive, cars are rapidly seeing costs of electronic components pushing 5 to 7% of their market value.
Across the industry, there are real opportunities and innovative designs on the horizon thanks to shrinking chip geometries and architectures. However, these opportunities do not come without real challenges for the industrial and manufacturing sectors. For design engineers, the goals have not changed, but the challenges faced from the new components are many and the time between cycles is shortening. Ever-smaller geometries and 3D-stacking hold promise but must be carefully tested and balanced to address the concerns for long-term reliability and quality when these solutions must then go into Design for Manufacturing (DFM) and production phases.