Confidence within the semiconductor {industry}’s capability to satisfy demand is being hindered by geopolitical tensions, worldwide commerce restrictions and the push for sovereignty, in response to a report from Capgemini.
Authors of The semiconductor {industry} within the AI period report warn that as nations compete for management over very important applied sciences and assets, geopolitical tensions proceed to affect the worldwide semiconductor provide chain.
As an illustration, the stream of parts, supplies and accomplished semiconductor merchandise has been hindered by worldwide commerce disputes, export restrictions and tariffs. For instance, Capgemini estimated that 14 client provide networks that depend on TSMC may very well be significantly disrupted by any navy escalation involving China and Taiwan. Deteriorating US-China ties have additionally given rise to setbacks within the type of prohibitions on sure merchandise and extra stringent controls.
The report, primarily based on a survey of 250 semiconductor corporations and 800 “downstream” organisations that use semiconductors in their very own merchandise, discovered that 58% of semiconductor organisations count on greater demand for neural processing models to accompany progress in generative AI (GenAI) adoption. The research additionally reveals that 57% of the chip producers surveyed anticipate elevated want for high-performance chips, and 56% count on to see larger demand for memory-intensive chips, which, in response to Capgemini, alerts a shift in the direction of superior processing.
The report reveals there’s elevated demand for customized chips and customized software program optimised to run on these chips.
Jiani Zhang, government vice-president and chief software program officer at Capgemini Engineering, mentioned superior platforms and software program are vital differentiators within the semiconductor {industry}, driving effectivity and scalability in design, manufacturing and deployment.
“With the rising complexity of AI, IoT [internet of things] and edge computing functions, the flexibility to combine domain-specific software program with {hardware} accelerators will outline management,” she mentioned. “To remain aggressive, semiconductor gamers should embrace co-optimisation throughout the stack, from chip structure to software interfaces, making certain they will meet the escalating calls for of data-intensive, low-latency markets.”
Capgemini mentioned that whereas the demand for AI chips, customized silicon chips and memory-intensive chips is predicted to extend over the following 12 months, the semiconductor {industry} must capitalise on rising alternatives. These embody design and cutting-edge, sustainable fabrication strategies, in addition to funding in home sourcing and nearshoring to reinforce stability.
“GenAI is driving accelerated demand for chips, and semiconductor firms face growing calls for from clients who need extra personalised and software-centric experiences,” mentioned Brett Bonthron, international high-tech {industry} chief at Capgemini.
“The {industry} ought to see this as a chance to ramp-up manufacturing and undertake a ‘chip-to-industry’ strategy that helps a full-stack, ‘software-first’ set of capabilities,” he added.
“Funding in cutting-edge fabrication strategies and design processes powered by AI and GenAI will probably be key to satisfy the specialised wants of rising functions. Equally, it’s essential that the {industry} additional enhances sustainable manufacturing processes and makes use of superior safety to safeguard mental property.”
