This week brings together market, manufacturing, and investment signals shaping semiconductor execution in 2026.
This edition covers record 2025 industry sales, new R&D programs in specialty devices and materials, and metrology upgrades targeting faster yield ramp and tighter process control.
It also tracks policy and capex moves affecting advanced-node expansion and ecosystem development, while continued memory tightness and AI infrastructure demand drive pricing, capacity allocation, and data center silicon roadmaps.
1. Global semiconductor market hits a new high at $791.7B in 2025
Global semiconductor sales reached $791.7 billion in 2025, up 25.6% year on year. Q4 accelerated sharply, driven mainly by logic and memory demand tied to AI, cloud, and advanced computing.
Logic became the largest product category, while memory followed closely behind. Industry forecasts now point toward nearly $1 trillion in global sales in 2026, with AI, automotive, IoT, and future wireless technologies as the main drivers.
2. Lam Research and CEA-Leti extend collaboration on specialty technologies
Lam Research and CEA-Leti expanded their long-running R&D partnership to focus on specialty semiconductor technologies, including RF, photonics, sensors, MEMS, and optical interconnects.
The goal is to shorten the time from new materials and process ideas to working, validated devices — especially where energy efficiency and AI integration matter most.
3. Texas Instruments agrees to acquire Silicon Labs for $7.5B
Texas Instruments announced a $7.5B all-cash acquisition of Silicon Labs, expanding its embedded and wireless connectivity portfolio. The deal adds roughly 1,200 products and strengthens TI’s position in industrial and embedded markets.
TI plans to move much of Silicon Labs’ production into its own 300 mm fabs, targeting around $450M per year in synergies within three years after closing.
4. India shifts semiconductor policy toward full ecosystem build-out
India’s FY2026–27 budget signals a move from fab-focused incentives toward a broader semiconductor ecosystem strategy. This includes equipment, materials, supply chain resilience, and local IP development.
The budget also expands electronics manufacturing incentives and introduces process changes aimed at reducing friction for imports, tooling, and large-scale manufacturing operations.
5. Siemens acquires Canopus AI to strengthen metrology and inspection
Siemens acquired Canopus AI to add AI-driven metrology and inspection capabilities to its EDA and manufacturing software stack.
The acquisition supports better pattern fidelity, faster yield ramp, and stronger digital-twin modelling as chips continue to shrink and production complexity rises.
6. SoftBank’s SAIMEMORY and Intel align on next-generation memory
SoftBank subsidiary SAIMEMORY signed a collaboration with Intel to advance Z-Angle Memory, a new memory architecture targeting high bandwidth and low power for AI workloads.
Prototypes are planned by FY2028, with commercialization targeted for FY2029 — reflecting the long timelines typical for new memory technologies.
7. Cerebras raises $1B to scale wafer-scale AI systems
Cerebras closed a $1B Series H funding round at a $23B valuation to expand deployment of its wafer-scale AI systems.
Its Wafer Scale Engine targets high-throughput AI training and inference, with customers already active across multiple regions and deployment models.
8. Memory prices surge as AI demand tightens supply
TrendForce sharply raised its 1Q26 memory price forecasts, citing widening supply gaps driven by AI servers and data centers.
DRAM and NAND prices are now expected to rise at record levels quarter-on-quarter, with PC, server, mobile, and enterprise SSD segments all under pressure.
9. Technical University of Munich tapes out a 7 nm AI chip
The Technical University of Munich announced the EU’s first university-designed AI chip at the 7 nm class, based on RISC-V and aimed at edge computing.
The project links academic chip design, AI workloads, and future European manufacturing, with production referenced to planned capacity in Dresden from 2028 onward.
10. Report points to $17B TSMC plan for 3 nm production in Japan
Reports indicate TSMC may invest $17B to add 3 nm production in Japan, marking a shift toward more advanced nodes in the country.
The move aligns with Japan’s broader push for semiconductor supply security and complements TSMC’s existing plans in Taiwan and Arizona.
11. Mature-node capacity tightens as AI servers crowd fabs
Vanguard International Semiconductor reported tight capacity at mature nodes, driven by strong AI server demand alongside recovering automotive and industrial markets.
Utilization is rising, order visibility is improving, and mature-process chips are becoming a bottleneck — highlighting that not all AI demand sits at leading-edge nodes.
