Rand Technology’s CEO, Andrea Klein, explains how the crossover of the artificial intelligence gold rush and EV rollout could trigger the next supply chain crisis
As the world increasingly leans in on AI-driven solutions, demand for semiconductors is expected to skyrocket, creating a dynamic landscape for OEMs. Beneath the surface of progress lies a critical challenge—potential repercussions on component availability. While many may prognosticate about AI’s impact on the supply chain, there is no frame of reference. Companies need to be prepared.
While EV adoption was thought to be the catalyst ushering in next generation technology and future component shortages, it’s actually the expected early adoption of AI that will be 2024’s driving force.
The explosion of AI begins with the buildout of the AI cloud infrastructure then will continue into on-premises AI infrastructure buildout and finally to the ‘edge’, otherwise known as on-device. This demand for new hardware and refresh of the edge product will drive demand for traditional components, advanced chips and a variety of board-level components, particularly power products.
Currently, foundries are pivoting to increase production of advanced nodes. However, this realignment takes time, is expensive and does not consider demand growth for standard components needed to support advanced chips. This is manageable now because the market is soft and many companies have an oversupply of components resulting from past long-term agreements. However, early adoption of AI could change the semiconductor digestion rate. To date, current infrastructure is insufficient to support next generation technology. Once again there will be semiconductor constraints, with companies competing for product.
To understand impending constraints, let’s examine the intricacies of supply chains to identify the potential triggers. Supply chains can be visualized as a pyramid, with different layers. At the base are raw materials, which are vulnerable due to factors including global politics, mining rights and smaller foundries competing with TSMC for their fair share.
Moving up the pyramid, a mix of issues persist at the foundry, substrate and OSAT level, due to a lack of visibility into future needs, insufficient investment, labor shortages and de-risking of dependence on China.
The third layer involves the expansion of cloud infrastructure and networking capabilities, which are essential to support next generation products. Further, the crossover of AI and EV is anticipated to be problematic. It is unclear how quickly this will happen but it will lead to a massive need for electronic components, setting in motion a domino effect with each layer of the supply chain feeling pressure.
This should serve as a wake-up call to hardware technology company CEOs, chief procurement officers and those in R&D departments to prepare for this renaissance or revolution in hardware technology. Companies need to build as much flexibility as possible in their bills- of-materials, strengthen their company’s muscles in commodity management and supply chain, heighten their supplier relationships and prepare for the future.