Year 2021 brought number of challenges within the semiconductor industry and components supply chain challenges. Murray Slovick summarizes the trend and looking ahead for 2022 component market in his article published by TTI Market Eye.
The ongoing supply crunch for chips has hurt production across a range of industries from cars to personal computers and smartphones. But while things should improve in the second half of 2022, the first six months could still see pockets of shortage across the industry, exacerbated by the moves of tech giants like Apple, Amazon, Meta (formerly Facebook) and Tesla – all of whom will start to bring certain elements of chip development in-house.
In all, 169 industries have faced the crunch, according to data gathered by Goldman Sachs. As a result, the current shortages are likely to persist into 2022, although even the hard-hit auto industry is showing signs that the worst is over and semiconductor supplies will resume at normal levels during the next fiscal year.
But while you can’t always trust statistics – a wise man once said they are used in an argument much the way that a drunkard uses a lamppost: for support instead of illumination – it is tough to dispute the numbers. They underpin the notion that despite all of the drama, the semiconductor industry has demonstrated great resiliency in the face of disruptive global challenges, including the chip shortage and the ongoing pandemic.
According to news reports citing SEMI’s “Worldwide Semiconductor Equipment Market Statistics Report,” global semiconductor equipment billings increased a robust 38 percent year-over-year to US$26.8 billion in the third quarter of 2021, an 8 percent rise from the prior quarter to register their fifth consecutive quarter-over-quarter record high. Ajit Manocha, SEMI president and CEO, was quoted as saying: “Strong secular demand for chips across a wide range of markets including communications, computing, healthcare, online services and automotive has fueled this tremendous run of record quarterly growth for semiconductor equipment.”
So, where do we go from here? A number of segments will provide the industry with abundant opportunities. Let’s take a look at a few of these.
Automotive: EVs Lead the Charge
In the automotive sector, the adoption of safety-related electronics systems has grown explosively. The automobile industry accounts for about one-tenth of total semiconductor demand, according to figures from SIA and Gartner. Electronic systems could account for half a car’s cost by 2030, up from about a fifth in 2000, according to Deloitte. Once pandemic lockdowns were eased and the economy started to recover, car sales picked up rapidly. All this has led to greater demand for chips.
Toss in the fact that the leading suppliers of chips to the auto industry is benefiting from a tailwind as more carmakers shift to electric vehicles. As such the number of passives like multilayer ceramic capacitors required for the proper operation of EV functions is also getting a boost.
It’s not just silicone chips that the increased demand is benefitting. The automotive industry’s requirement for 6-inch silicon carbide wafers is expected to reach 1.69 million units in 2025 thanks to the rising penetration rate of EVs and the trend towards high-voltage 800V EV architecture, according to a report from TrendForce.
Artificial Intelligence: “AI Everywhere”
Not long ago, “smart” really just meant “connected,” such as smartphones, smart TVs and other devices connected to the internet. Today, “smart” increasingly means powered by artificial intelligence (AI) – generally machine learning algorithms – and capable of helping designers in increasingly innovative ways.
The recent advent of machine learning (ML) and the increasing complexity of EDA tasks have aroused more interest in incorporating AI to solve design tasks. As an example, consider that smartphones now use AI algorithms to do everything from maintaining call quality to helping take better pictures. AI also helps in automating tasks ranging from data collection, communications and robotics/factory automation to EDA in chip design.
Many of these AI platforms will increasingly require high-speed connectivity. You can certainly expect that in 2022 you’ll be using more AI-enhanced tools to help you innovate.
5G: Becoming Familiar with mmWave
Turning 5G wireless communication into a widespread reality will require the use of mmWave frequencies. It’s critical to get a handle on the essentials involved in working in this range.
First and foremost, the rules and specifications (such as those defined by the 3GPP Specification) are different for mmWave. This specification defines the frequency ranges where 5G can operate: 4.1 GHz to 7.125 GHz and 24.25 GHz to 52.6 GHz.
There are also a variety of different RF design challenges in mmWave, largely because as frequency increases, wavelength decreases. For one thing, different filtering technologies are needed. For another, 5G technology also may require numerous antennas because it uses higher-frequency radio waves with short ranges. Typically, several radio-frequency integrated circuits (RFICs) have to be fitted in order to emit radio waves in multiple directions – or, alternately, a design must be able to deliver a secure mmWave wireless communication environment in two directions using just a single RFIC.
Moore’s Law: Not Dead Yet
As Thomas Jefferson so eloquently noted, some truths are self-evident – that is, so obviously correct that to attempt rebuttal is to invite ridicule. Most today would agree that a 2022 list of incontrovertible truths would have to include finding a viable alternative to Moore’s Law, which stems from Gordon Moore’s observation in 1965 that the number of transistors on a processor would double every two years.
As the die shrinks to approach three nanometers and beyond (TSMC has just kicked off pilot production of chips built using 3nm process technology at its Fab 18 in southern Taiwan, and is expected to move the process to volume production by the fourth quarter of 2022), industry is reaching the physical transistor size limit.
Of course, “more than Moore” (the catchy phrase used to describe other ways of scaling systems beyond purely dimensional scaling of the silicon) is basically about advanced packaging. A new approach to IC design subdivides a system into functional circuit blocks called “chiplets,” in which a single chip is broken down into multiple smaller independent constituents which make up a chip built out of multiple smaller dies connected together and combined in new ways.
Another approach – the 3D stacking of integrated circuits – also has potential technical benefits and could help to current chip shortage concerns. This solution stacks chips on top of each other, which enables the use of a shorter wires, reducing power requirements and making the timing paths smaller, allowing better performance.
Quantum Computing is Getting Closer
For decades, scientists have hypothesized about computers based on the mathematics and physics used by subatomic particles. Quantum computers employ qubits to represent information in quantum form.
Unlike data represented by a traditional computer that operates in bits, quantum data may be both a “1” or “0” simultaneously. In theory, quantum computers can perform certain calculations much faster and more efficiently than digital computers. IBM and other companies such as Google and Microsoft are working to scale towards what is called the “quantum advantage,” where quantum computers are either cheaper, faster or more accurate than classical computing.
IBM used its annual Quantum Summit in November to unveil the 127-qubit Eagle processor. As outlined in its hardware roadmap, IBM aims to unveil its 433-qubit Osprey processor next year. Condor, a 1,121-qubit processor due in 2023, would enable tasks like error correction.
Welcome to the Metaverse
In 2022, you’ll hear more and more about the concept of a metaverse, digital worlds that exist in parallel with the physical world we live in. The term “metaverse” first appeared in Neal Stephenson’s 1992 sci-fi novel Snow Crash where it referred to a 3D virtual world inhabited by avatars of real people. It has also been used by other science fiction novelists such as Ernest Cline, who depicted a version of the metaverse in his book and the subsequent 2018 movie adaption, “Ready Player One.”
Some foresee the metaverse as a successor of our current internet, with characteristics that include unprecedented interoperability; users will be able to take their avatars and goods from one place in the metaverse to another. Anywhere you go, you can just bring that world with you wherever you want. Inside these metaverses we will carry out many of the functions we’re used to doing in the real world, including working, playing, and buying or selling things.
Focus on Reliability
Reliability will be among the discussion items that will influence passive component design and selection next year. This will require a better understanding of the component failure mechanisms across the entire operating spectrum of temperature, vibration and other parameters.
Where there is a lack of standards to enable successfully extending the reliability of electronic components, users will have to adapt a data-based approach to determine the extended use of these components. Where standards do exist, such as developed by the Automotive Electronics Council (AEC) Component Technical Committee (which endorses the AECQ-200 standard as the guideline for validating automotive grade passive electronic components), newly-introduced components will have to be designed to be compliant with these standard requirements – even if, in this case, the end application is not necessary related to the automotive segment.