- Broadcom and TSMC have pointed to growing bottlenecks in AI chip supply chains.
- The issue reflects surging demand for advanced semiconductors used in data centres and AI workloads.
What happened: Strain on chip manufacturing capacity
Broadcom and TSMC have flagged increasing pressure on the supply chain for artificial intelligence chips, as demand continues to outpace manufacturing capacity.
According to Capacity Media, the companies highlighted constraints affecting the production of advanced semiconductors used in AI systems, particularly those deployed in large-scale data centres.
The surge in demand is being driven by rapid adoption of generative AI and machine learning applications, which require specialised chips capable of handling high-performance computing workloads.
TSMC, the world’s largest contract chip manufacturer, plays a central role in producing advanced semiconductors for many leading technology companies. Broadcom, meanwhile, supplies chips and infrastructure technology used in networking and data centre environments.
A key bottleneck is TSMC’s Chip-on-Wafer-on-Substrate (CoWoS) advanced packaging capacity, which is critical for assembling high-performance AI chips. The technology is currently fully booked, with new capacity not expected to ease constraints until late 2026. Broadcom CEO Hock Tan noted during a recent earnings call that demand visibility remains strong through 2026, but cautioned that packaging limitations are creating delivery delays.
The report indicates that supply chain challenges include limited manufacturing capacity, complex production processes and the need for advanced fabrication technologies.
As AI adoption accelerates, these constraints are becoming more visible, raising concerns about whether supply can keep pace with demand.
Also Read: US grants TSMC licence for chipmaking tools to China plant amid export control tweaks
Also Read: Nvidia asks TSMC to boost H200 chip output as Chinese demand surges
Why it’s important
The strain on AI chip supply chains highlights a critical bottleneck in the development of artificial intelligence infrastructure.
While software innovation has advanced rapidly, the availability of hardware remains a limiting factor. Without sufficient supply of advanced chips, companies may struggle to scale AI systems and services.
For cloud providers and data centre operators, access to high-performance semiconductors is essential to supporting AI workloads.
From a financial perspective, supply constraints can drive up costs and influence investment decisions across the technology sector. Companies may need to secure long-term supply agreements or invest in alternative solutions to mitigate risk.
The situation also underscores the strategic importance of semiconductor manufacturing in the global economy.
Governments and companies alike are increasingly focused on strengthening supply chains and reducing dependence on single sources of production.
The involvement of companies such as Broadcom and TSMC illustrates how central these issues have become to the technology industry.
As demand for AI continues to grow, resolving supply chain constraints will be key to sustaining momentum in the sector.
The current pressures therefore point to a broader reality: the future of artificial intelligence is closely tied to the capacity and resilience of the global semiconductor supply chain.
