PwC Report: $1.5 Trillion Semiconductor Buildout Fueling AI Data Centers, EVs, and a 100,000-Engineer Talent Race
AI and EVs drive record chip demand; 2024-2030 fab spend tops $1.5T as CHIPS Act accelerates buildout. Plan for AI accelerators, SiC/GaN, HBM, and 100k new engineers.
How Investment in Semiconductors Is Driving AI and EVs
AI workloads and electric vehicle adoption are pushing demand for advanced chips to new highs. PwC projects global fab investments will exceed US$1.5tn from 2024 to 2030-on par with the past two decades combined. Policy support is accelerating the buildout, with the US CHIPS Act catalyzing large, multi-year projects across the supply chain.
For IT, finance, and product teams, this is a signal: capacity, cost curves, and roadmaps are shifting. The companies that plan for this cycle now will have an operational edge by mid-decade.
Where the money is going
SEMI expects 18 new fab construction starts in 2025, with most coming online in 2026-2027. The US CHIPS Act has helped trigger more than US$630bn in announced supply-chain investments across 28 states, stacking private capital on top of federal incentives.
Texas Instruments plans US$60bn across seven US fabs-the largest foundational manufacturing commitment in US history. Intel is deploying more than US$100bn across four states, supported by CHIPS funding that includes US$7.86bn in direct support and US$3bn for secure government manufacturing opportunities. For the source analysis, see PwC's Semiconductor & Beyond 2026 report and CHIPS for America program details.
PwC: Semiconductor & Beyond
US CHIPS for America
AI data centers are now the demand anchor
PwC forecasts the global server market will exceed US$300bn by 2030, making server and network the largest chip demand segment. AI accelerators are set to drive ~50% of that segment's revenue by 2030 as inference and training scale out.
Hyperscalers are building custom accelerators to control cost, supply, and performance. High-bandwidth memory is surging alongside, as workloads require faster data movement for modern AI models.
EVs are rewiring vehicle electronics
Automotive is the second-fastest growing segment, with a 10.7% annual growth rate projected by PwC. Hybrids and EVs could reach 50% of sales by 2030, turning cars into software-defined, sensor-rich systems.
Silicon-carbide (SiC) and gallium-nitride (GaN) devices are set to replace legacy silicon across traction inverters and fast-charging hardware due to better efficiency and thermal performance. By 2030, more than 60% of traction and charging semiconductors in vehicles are expected to be SiC or GaN, up from a combined 23% today.
Consumer and edge keep the flywheel turning
Home appliances are projected to grow 5.6% by 2030, while AR/VR and personal robots are projected at 24.5% and 12.9%, respectively. AI appliances-from smart fridges to autonomous vacuums-need local compute to cut latency and improve privacy.
That drives demand for AI-grade NPUs, efficient microcontrollers, PMICs, and connectivity chipsets at the edge. Expect broader use of on-device inference and memory-rich designs in everyday products.
What leaders should do next
- IT & Development: Plan for multi-accelerator stacks (Nvidia, AMD, and custom silicon). Build portability into models and tooling. Track HBM availability as a gating factor for both cost and throughput.
- Finance: Model longer depreciation cycles for fabs and data center hardware. Expect capex front-loading in 2025-2027 with utilization gains into 2028-2030.
- Product: Qualify SiC/GaN suppliers early and design for multi-sourcing. For edge devices, prioritize thermal budgets and memory bandwidth to keep on-device AI responsive.
- Talent: The industry will need an additional 100,000 engineers by 2030. Upskill teams on AI hardware, embedded systems, and data center architectures.
If you're building AI capability across roles, explore focused upskilling paths here: AI courses by job.
Key facts
- Global semiconductor fab investments from 2024-2030 will exceed US$1.5tn.
- AI accelerators are projected to reach 50% of data center semiconductor revenue by 2030.
- The industry requires an additional 100,000 engineers by 2030.
Bottom line
AI, EVs, and edge devices are setting the pace for chip demand through 2030. The capacity being built now will define pricing, availability, and performance tiers for years. Align your sourcing, architectures, and hiring with this cycle while incentives and supply pipelines are still being set.
