Access Official Technical PDF

Share Summary:LinkedInEmail

Integrated circuit scaling is more than a story about fitting additional components onto silicon. It changes the economics of electronic functions, influences how systems are designed, and creates tradeoffs in manufacturing yield, heat, and interconnection. For industrial automation engineers and technical decision-makers, these principles help explain why integration can simplify equipment—and why a larger chip is not automatically the best system choice.

Integrated circuit scaling changes the economics of electronics

The central economic idea is that a semiconductor structure can contain more components without requiring a proportional increase in its basic material. For simple circuits, adding components to an integrated circuit can therefore reduce the cost per component. Integration can also reduce the effort involved in assembling and designing a system by making complete circuit functions available as units.

That cost relationship has limits. As circuit complexity rises, manufacturing yield can fall, offsetting the savings from placing more components on each chip. The lowest-cost circuit size is therefore not fixed: it depends on the technology and manufacturing conditions at a given time. Moore’s 1965 analysis described the expected minimum-cost complexity as rising as the overall cost curve fell.

This distinction matters in system planning. The objective is not simply to maximize component count on a chip. It is to minimize the total cost of delivering the required system function, including the engineering needed to develop and produce it.

Reliability and system architecture

Integration can improve system-level reliability by replacing multiple separately assembled parts and connections with a functional circuit unit. The original analysis pointed to integrated electronics as valuable in applications where size, weight, and reliability were important, and noted that integrated circuits had demonstrated high reliability even while production volumes remained relatively low.

For automation systems, the broader design lesson is to evaluate reliability at the level of the complete function, not just the individual component. A highly integrated device may reduce the number of separate elements that must be connected, but its usefulness still depends on whether its function, availability, and production economics suit the application.

Nor does integration require every system function to reside on one chip. Large systems can be built from smaller, separately packaged functions connected together. Alternatively, more functions can be combined into larger arrays. The appropriate choice depends on engineering costs, production quantities, and the value of flexibility. Reusable functional blocks can help avoid concentrating disproportionate design effort in a single large array.

Designing for density: interconnection and yield

Increasing component density requires more than shrinking individual devices. Connections between components take space, so interconnection patterns can limit how much circuitry fits into a given area. The 1965 paper described multilayer metalization separated by dielectric films as one approach to creating more complex structures while reducing wasted area for connections.

Manufacturing yield is another key consideration. A larger circuit can be economically attractive only if enough completed structures meet requirements. The source argued that yields could improve through engineering effort and that the same general pattern seen in smaller integrated circuits might support larger arrays. That is a forecast about the potential of process improvement, not a guarantee that any particular complex design will be economical.

When assessing a highly integrated option, decision-makers should therefore consider:

  • Function: Does the circuit provide the combination of functions the system needs?
  • Manufacturing economics: Do expected costs and yields support the intended production volume?
  • Interconnection: Can the required connections be implemented without undermining the benefits of density?
  • Engineering effort: Can design work be reused or automated, or will the array require disproportionate development effort?

Thermal design and the limits of integration

Concentrating more components raises an immediate question: how will the circuit dissipate heat? The 1965 analysis noted that integrated structures are two-dimensional, leaving a surface near the points where heat is generated. It also argued that power demand is closely related to driving circuit lines and capacitances, which can be limited when a function occupies a small area.

These observations do not remove thermal design from consideration. They show why circuit dimensions, layout, power, and the area of the integrated function need to be considered together. In system design, increasing density should be evaluated alongside the heat generated and the means available to remove it.

Digital and linear circuits scale differently

Integration can reshape digital systems substantially, but linear circuitry faces different constraints. Large-value capacitors and inductors are difficult to integrate because energy storage requires volume. The source anticipated that these components would remain part of linear designs, even as other functions became integrated.

This points to a hybrid approach: integrate stages or functions where doing so is practical, while retaining larger tuning or energy-storage elements where the circuit requires them. The same principle applies to system architecture more broadly. Integration is a design tool, not a requirement to put every element into a single package.

Moore’s Law: a forecast and an industry driver

Moore’s original 1965 projection described the observed pace of increasing component complexity and extrapolated it forward. In a later interview included with the source material, Moore explained that the original estimate was roughly a doubling each year; in 1975, he revised the expectation to about every two years. The commonly repeated 18-month figure was not his original statement, but a later formulation associated with increased computer performance.

Moore also described how a projection became a goal that influenced industry planning. For automation leaders, the useful takeaway is not to treat a historical trend as a guaranteed schedule for every technology. It is to recognize that expectations about component density can shape road maps, design choices, and investment decisions.

Practical implications for automation teams

Integrated circuit scaling can lower component costs, support compact functional units, and simplify some aspects of system design. Its benefits depend on balancing circuit complexity against yield, interconnection, thermal demands, and engineering cost. Linear functions may also retain components that are not readily integrated.

When comparing architectures, automation teams should start with the required system function and evaluate the complete implementation—not just the number of components per chip. Choosing between a large integrated array and interconnected smaller functions is an economic and engineering decision. A clear view of those tradeoffs helps plant and system design teams use integration where it delivers practical value without assuming that maximum density is always the right answer.

Access Official Technical PDF

About The Author