A 6-point action plan for software companies to manage the global chip shortage crisis
Since mid-2020, several global factors including a drought in Taiwan and the COVID-19 pandemic have resulted in a semiconductor chip shortage, which worsened further as demand for electronics shot up during the pandemic.
As offices and schools moved to homes and digital services became essential, in 2020 alone, 297 million PCs were sold, up 11% over the previous year. Such market spikes resulted in a 17% increase in semiconductor demand in 2021 versus 2019, which thinned inventory to as few as five days’ worth compared to the usual 50 days. That is why this situation is also referred to as a supply shock in the semiconductor value chain.
Vital electronics such as 5G networks, cloud computing devices, automobiles, gaming consoles, and medical devices depend on electronic chips with component densities ranging from 800nm to the tightly packed 5nm chips used in high-performance equipment. Only a few companies around the world produce a majority of chips, led by Taiwan-based TSMC and followed by Samsung, Intel, and others.
As semiconductor manufacturing is a complex process that is time-intensive, costly, and labor-intensive, the chip shortage crisis is likely to continue till 2023 or 2024, until the hundreds of billions of dollars poured into expansions start to bridge the gap. While some manufacturing companies have coped by salvaging parts and reducing functionalities to ship faster, many have stalled production for want of essential chips, whose inventory levels have plummeted, and due to prices of components going up by more than tenfold.
Though chipmakers have begun to accelerate their investments to open new semiconductor fabrication centers, the world of software development has to look inward and work towards application performance optimization. To do more with less. Better code can compensate for hardware inadequacies by squeezing more computing power from existing chipsets, which comes from optimizing applications and minimizing data and processing requirements.
In his earnings call in July 2021, Elon Musk spoke of how Tesla managed to swap substitute chips after rewriting its firmware code to cope with the supply gap, setting an example. Until the supply chain eases, and for even longer, there is a need for drastic course correction in the software development industry.
Here is a six-point action plan for software companies to deal with the ongoing global chip shortage crisis by focusing on optimizing application performance.
- Move to and optimize for the cloud: Offload customized code onto cloud servers if your on-premises IT infrastructure struggles with an increased workload. Use AI algorithms to optimize code and reduce cloud spend.
- Unify and minimalize your code: Minimalistic, elegant, lean coding practices increase efficiency and cut downtime and repair costs. Simplified software with lean code helps increase flexibility in product development. Unify your technology stack and clean up unnecessary processes. Simplify UI or UX elements to favor utilitarian functioning. Become essentialists. Steer clear of feature bloat by offering only the essential features as standard, moving the rest to opt-in features.
- Understand the root cause; don't delay the inevitable: It is natural to add computing resources through capacity expansion as a stopgap to avoid lags in your products and services. However, this is only going to be a short-lived scalability fix as it does not address the real root cause: the lack of programming efficiency. Analyze your tech stack and solve your performance problems through efficient, lean coding.
- Switch to austerity mode: Software developers have been building larger and more feature-rich offerings with powerful functionalities, assuming that hardware innovation will keep pace with growth. Facing the reality, developers should now change their perspective and work towards optimizing applications through stringent planning, execution, monitoring, and optimization.
- Adopt the SaaS mindset: As you expand or optimize, choose to build from time-tested, easy-to-deploy integrations, with API support, serverless computing, and containerization possibilities.
- Leverage application performance monitoring: DevOps teams can choose full-stack monitoring solutions that give end-to-end observability and powerful insights into application performance. A good APM solution helps track real-time application behavior, optimize response time, understand and adjust external dependencies, monitor custom integrations, and even deploy AI-powered alerts.
In summary, software companies should adopt a comprehensive monitoring approach to be on top of their game—optimize their applications to run with current chips for longer, reduce bloatware, eliminate redundancies, go for lean code, and shift to the cloud whenever possible.
Though the above approach seems to be the way forward until the global semiconductor supply chain expansion materializes, better coding and holistic monitoring with a minimalistic, frugal mindset will also prove to be a winning strategy for years to come.