bottleneckScore 80/100Research
NVIDIA Data Center supply chain constrained by networking components, switches, NICs, and cables—not just GPU chips
Jensen Huang· NVIDIA· AI· 2022-08-25· about NVIDIA (NVDA)
“We were challenged this quarter with a fair amount of supply chain challenges because as you know, we don't just sell the GPU chip, but these systems are really complex, with a large number of chips in the system components that we offer, like HGX... these data centers sit idle until the last piece comes together. And the last piece includes very complicated switches and very complicated NICs and networkings and cables.”
Why it matters
NVIDIA's data center systems are bottlenecked not by GPU production but by networking infrastructure (switches, NICs, cables), which suggests suppliers of networking components and data center interconnect solutions are the true constraint in scaling AI infrastructure.
Investment implication
Suppliers of high-speed Ethernet switches, network interface cards, and data center cabling are critical to AI capex cycles. Companies like Broadcom (networking chips), Arista Networks, or specialized networking suppliers become picks-and-shovels plays for AI infrastructure buildout.