This week’s semiconductor updates highlight record-breaking valuations, sovereign AI infrastructure projects, and major technology partnerships shaping the next era of advanced computing and quantum innovation.
1. Global Silicon Wafer Market Set for 5.4% Growth in 2025, Record Output Projected by 2028
AI-Driven Demand Accelerates Silicon Wafer Recovery
Global silicon wafer shipments are forecast to rise by 5.4% in 2025, reaching 12,824 million square inches (MSI), according to SEMI’s annual report. The recovery follows a period of cyclical slowdown, supported by rapid expansion in AI-related demand, particularly for advanced epitaxial wafers used in cutting-edge logic devices and polished wafers supporting high-bandwidth memory (HBM). Growth in non-AI applications is progressing gradually but is expected to strengthen as macroeconomic conditions improve.
Industry Collaboration and Long-Term Outlook to 2028
SEMI projects wafer shipments to reach 15,485 MSI by 2028, setting a new global record. This projection reflects robust input from members of the Silicon Manufacturers Group (SMG), part of SEMI’s Electronic Materials Group (EMG). The forecast encompasses both polished and epitaxial wafers, excluding reclaimed and non-polished types. With AI computation expanding across data centers and edge systems, silicon wafers remain the indispensable substrate for semiconductor device fabrication worldwide.
2. U.S.–Asia Trade Diplomacy Redefines Strategic Equilibrium Between Washington and Beijing
Economic Negotiations Shape Supply Chain and Tariff Policy
During a five-day diplomatic tour through Malaysia, Japan, South Korea, and China, U.S. President Donald Trump secured new trade agreements and access to critical minerals while seeking reduced tariffs for allied nations. The trip included Japan’s $550 billion investment commitment in the United States and new Southeast Asian agreements on border and trade cooperation.
U.S.–China Dialogue Highlights High-Stakes Interdependence
The concluding meeting with President Xi Jinping in Busan marked a turning point in the trade dialogue between the world’s two largest economies. Both sides agreed to de-escalate tariffs and reopen agricultural imports. This pragmatic recalibration aims to stabilize global supply chains for semiconductors, rare earths, and energy products.
3. South Korea and NVIDIA Build AI Infrastructure Hub with 260,000 GPUs
National AI Factories and Sovereign Compute Expansion
The Korean Ministry of Science and ICT, together with NVIDIA and major partners including Samsung, SK Group, Hyundai, and NAVER Cloud, is deploying over 260,000 NVIDIA GPUs to build sovereign AI infrastructure. Backed by a $3 billion investment, the plan positions Korea as a top-three AI powerhouse.
AI Ecosystem Expansion and Quantum Research Integration
NVIDIA and Korean institutions such as KISTI are advancing sovereign foundation models and quantum research using NVIDIA accelerated computing. This initiative integrates AI-RAN, 6G, and startup incubation, placing South Korea at the forefront of AI-driven industrial transformation.
4. NVIDIA and Oracle to Deliver DOE’s Largest AI Supercomputer for Scientific Discovery
100,000 Blackwell GPUs to Power U.S. Research Acceleration
The U.S. Department of Energy, NVIDIA, and Oracle are collaborating to build “Solstice,” the nation’s largest AI supercomputer at Argonne National Laboratory, featuring 100,000 NVIDIA Blackwell GPUs. A companion system, “Equinox,” adds 10,000 more GPUs.
Public–Private Model to Strengthen U.S. AI Leadership
Developed under a public–private framework, these systems will power frontier scientific models and AI workflows across physics, materials science, and healthcare, reinforcing U.S. leadership in AI-powered research.
5. NVIDIA and Uber Partner to Scale Global Level-4 Autonomous Mobility Network
Hyperion 10 Platform Powers Next-Generation Robotaxi Deployment
NVIDIA and Uber are collaborating to deploy a global Level-4 autonomous mobility network using NVIDIA DRIVE AGX Hyperion 10. Uber’s rollout targets 100,000 vehicles by 2027.
Expanding the Level-4 Ecosystem and Safety Certification Standards
Automakers such as Stellantis, Lucid, and Mercedes-Benz are joining the ecosystem, while NVIDIA introduces Halos Certified — the first AI safety framework for autonomous fleets.
6. NVIDIA and Nokia Launch AI-RAN Platform to Accelerate Transition from 5G to 6G
$1 Billion NVIDIA Investment Strengthens Telecommunications Leadership
NVIDIA and Nokia will co-develop AI-RAN systems using the new Aerial RAN Computer (ARC-Pro), a 6G-ready compute platform integrating edge AI and connectivity. NVIDIA will invest $1 billion in Nokia to drive the initiative.
AI-Native Networks Enable Edge Intelligence and 6G Readiness
The collaboration unifies radio and AI workloads on an accelerated infrastructure, simplifying upgrades from 5G to 6G and boosting spectral efficiency and energy optimisation.
7. NVIDIA Introduces NVQLink to Bridge Quantum and GPU Supercomputing
Hybrid Quantum–Classical Systems Enable Next-Generation Research
At GTC Washington D.C., NVIDIA unveiled NVQLink — an interconnect architecture linking quantum processors with GPU supercomputers.
U.S. National Labs Pioneer Accelerated Quantum Supercomputing
Developed with the U.S. Department of Energy, NVQLink will integrate into CUDA-Q for unified programming across CPUs, GPUs, and quantum systems, enabling new breakthroughs in hybrid computing.
8. NVIDIA Reaches $5 Trillion Valuation Amid Record AI Demand and Strategic Expansion
Market Capitalization Surpasses Global GDP Benchmarks
NVIDIA’s valuation surpassed $5 trillion, driven by massive demand for AI GPUs and expansion into sovereign AI and 6G infrastructure.
AI Partnerships and Investments Reinforce Growth Outlook
The company continues to partner globally — including with the DOE, Nokia, and Uber — reinforcing its dominance in the AI-accelerated computing era.
9. onsemi Launches Vertical GaN Power Devices to Drive AI and Electrification
Breakthrough vGaN Technology Enables Higher Efficiency
onsemi introduced its vertical gallium nitride (vGaN) power platform, achieving over 1,200V operation and halving energy loss.
Expanding Energy Efficiency for High-Demand Applications
The technology supports miniaturisation and higher efficiency for AI servers, EVs, and renewable systems — reducing size and cooling needs.
10. GlobalFoundries Invests €1.1 Billion to Expand Chip Manufacturing in Dresden
SPRINT Project Strengthens European Semiconductor Sovereignty
GlobalFoundries will expand its Dresden fab under Project SPRINT, surpassing one million wafers per year capacity by 2028, supported by German and EU funding.
European Semiconductor Cluster Gains Global Momentum
The expansion cements Silicon Saxony as Europe’s leading semiconductor hub and strengthens Europe’s strategic autonomy.
11. NXP Expands Automotive and Edge AI Portfolio Through Acquisitions of Aviva Links and Kinara
Advancing In-Vehicle Connectivity and Edge Intelligence
NXP completed the acquisitions of Aviva Links and Kinara to enhance automotive SerDes and AI edge processing.
Integration Strengthens NXP’s AI-Driven Product Roadmap
The acquisitions enable faster adoption of ADAS and embedded AI, reinforcing NXP’s leadership in automotive and industrial systems.
12. AWS Activates Project Rainier: Half a Million Trainium2 Chips Online for AI Compute
Anthropic Partnership Scales One of the World’s Largest AI Clusters
AWS launched Project Rainier with nearly 500,000 Trainium2 chips, providing fivefold the compute power used for Claude model training.
Energy Efficiency and Sustainability at Hyperscale
Powered entirely by renewable energy, Rainier achieves industry-leading water efficiency, advancing AWS’s net-zero carbon goals.
13. Keysight Launches Quantum System Analysis to Accelerate Superconducting Qubit Design
Unified Simulation Reduces Cryogenic Test Cycles
Keysight’s Quantum System Analysis enables end-to-end simulation of superconducting qubit systems, cutting iteration time and cost.
Higher Fidelity Through Time Dynamics and Fridge Line Optimisation
The platform enhances qubit coherence by modelling noise and thermal photon leakage — extending to spin and hybrid modalities.
14. HUMAIN and Qualcomm Partner to Deploy Edge-to-Cloud AI Infrastructure in Saudi Arabia
200 MW Inference Capacity Targets Global Workloads from 2026
HUMAIN and Qualcomm will deploy AI200 and AI250 racks delivering 200 MW of inference capacity, enabling sovereign AI infrastructure in Saudi Arabia.
Hybrid AI Stack and Sector Solutions for Sovereign Capability
The collaboration combines Qualcomm accelerators and HUMAIN infrastructure to support vertical AI solutions across multiple industries.
15. Cadence Q3 2025: Record $7.0B Backlog and Revenue Outlook Raised
Strong Quarter Across IP and Hardware; Margin Expansion Continues
Cadence posted $1.339B in Q3 revenue and raised full-year guidance following record backlog and strong AI hardware demand.
Portfolio Expansion and M&A Momentum
Cadence is acquiring Hexagon D&E to enhance structural and system analysis, strengthening its AI-driven EDA ecosystem.
16. SK hynix Q3 2025: Record Revenue and Profit on AI Server Memory Demand
HBM and High-Capacity DDR5 Drive Historic Performance
SK hynix reported KRW 24.45T in revenue and a 52% net profit margin, driven by AI server memory and HBM3E adoption.
HBM4 Ramp and Capacity Expansion Support 2026 Demand
HBM4 development is complete, with mass production starting in Q4 2025, ensuring leadership in next-gen AI memory supply.
