Steering supply chains to manage mounting complexity in the semiconductor industry

Many of these pressures were nothing new, but the ripple effects of the pandemic, worsened by an ongoing trade war between the US and China, plunged supply chains into deeper uncertainty.

Here, authors Brent Wilson, Senior Vice President, Global Supply Chain at Qualcomm and Susan Gosnell, Senior Manager, Industry, at Kinaxis, discuss the supply chain challenges in the semiconductor industry.

The shortage has largely eased in 2023. The COVID-related demand for PCs, smartphones, tablets, and consumer electronics has alleviated and it’s likely that non-consumer electronics markets will follow suit in 2023 creating excess supply, however the semiconductor supply chain is no less complex. Even with semiconductors becoming more available, there are still inventory imbalances with a surplus of some chips and unavailability of others which is typical for the industry.

Semiconductor manufacturing is a multi-faceted process that requires coordination of different stages including design, assembly, testing and packaging, and sourcing of various parts and materials. Issues in one part of the supply chain can cause ripple effects through the entire system. In addition, there are differing chip requirements, either due to government regulations or different products being produced in different countries. So, how can organisations navigate these mounting challenges and move forward with confidence?

Embracing concurrent planning

Semiconductor chips can be tailored to suit different markets and applications. While ideal for customer requirements, this adds complexity due to the number of use cases being supported. For example, when we look at cellular devices, chips that connect with basebands can vary to meet different functionalities, such as radio frequency connectivity or power requirements. This creates a need to balance a number of competing resources across a variety of customised chipsets.

From a single, concurrent supply chain planning platform, manufacturers can quickly see the full picture of their supply chain end-to-end. In the semiconductor space, demand planning capabilities can improve forecast accuracy by leveraging inputs from multiple market signals, such as sales, marketing, finance, customer behaviour, and suppliers to improve planning accuracy with advanced forecasting models and machine learning. Businesses can be empowered to identify demand at risk and make decisions that are based on sound financial and impact modelling and will deliver on customer promises.

Although shortages are not posing the same risk to manufacturers now, it would be presumptuous to assume that supply will remain fully stable over the coming years. Raw materials, such as resins used in semiconductor assembly, came in short supply during the pandemic, and businesses should plan for the possibility of such scenarios again. Concurrent planning provides visibility into material and capacity constraints to help planners align demand with supply. 

Concurrent planning techniques can enable organisations to run multi-scenario events using various tactics such as adding capacity, run rates, prebuild, or balance across resources to devise the potential impact and tweak operations to suit. They can prepare in advance for potential shortages and source alternatives. In builds of semiconductors, this visibility can give businesses confidence to pace production in line with their corresponding baseband, for example.

Bringing everyone on to the same page

Concurrent planning allows organisations to gain full end-to-end visibility of the supply chain, but it also helps to break down siloes. Traditional, siloed planning, usually completed with disparate and non-collaborative technologies, means information is passed from one person to the next and usually at a pace that’s not suitable for meeting market changes quickly. Concurrent planning is akin to a virtual meeting, where every stakeholder is part of the same conversation, enabling agility, speed, and efficiency in decision making.

To further enhance the user experience, tailored metrics and data can provide easy-to-digest information for all users. Responsibilities can be delegated for faster, smarter collaboration, with AI and predictive algorithms able to automate routine tasks and escalate any exceptions. Advanced analytics can complement AI to improve accuracy and drive faster decisions.

Responsibility-based, context-driven, cross-departmental collaboration provides the real-time supply chain information that planning teams need to make decisions to solve problems or capitalise on opportunities.

Easing supply chain pressures

While the situation of shortages around semiconductors has improved following a post-pandemic return to normality, the industry can’t afford to fall behind when market conditions change in the future. The process of semiconductor design, all the way to product implementation, is one of complexity. To address potential supply chain volatility, businesses can make use of demand and supply planning technologies and leverage internal data inputs as well as outside-in external signals for accurate insights. Concurrent planning capabilities bring end-to-end visibility across the entire supply chain and provide decision intelligence that can be used to see the impact of changes to a plan while encouraging collaboration across departments and breaking down silos. Organisations are then primed to navigate any uncertain situations that may come their way.