New Taiwan adv. packaging site; M&A activity; GF’s CPO solution; Apple looks beyond TSMC; EU chip companies map; strong earnings; MIT’s new lidar chip; gallium production; AI inferencing at sea; quantum HW funding; PCIe 7 tool; home data centers; AI security warnings; temperature impact on EVs.
ASE and WUS are jointly building a ~$1.1B advanced packaging hub in Kaohsiung, Taiwan, for fan-out chip-on-substrate (FOCoS) and flip-chip ball grid array (FC BGA) technologies. The new site is expected to be completed by September 2029.
China is aiming to use 70% domestic chips by the end of this year, reports Nikkei Asia.
Apple reportedly held early-stage talks with Intel and Samsung about producing main device processors in the U.S., as the company explores additional domestic manufacturing options beyond TSMC, according to Reuters. Discussions point to continued pressure on major chip customers to diversify advanced-node supply chains amid capacity, geopolitical, and U.S. manufacturing concerns.
Siemens partnered with Xometry to develop an AI-native marketplace to bring together suppliers and buyers of custom manufacturing. Siemens also invested about $50M into Xometry, and noted it has invested $1B in U.S. manufacturing over the last 5 years.
Nvidia and IREN inked a deal to deploy up to 5 GW of Nvidia’s DSX-aligned AI infrastructure across IREN’s Sweetwater, Texas campus, with NVIDIA receiving rights to invest up to $2.1B.
Innovations
Nvidia, PulteGroup, and Span are putting mini data centers in homes to take advantage of unused local grid electrical capacity.
Panthalassa raised $140M to complete a pilot fab in Oregon and accelerate deployment of AI inferencing at sea using power generated from ocean waves.
ST expects to generate more than $3B in cumulative space-related semiconductor revenue from 2026 through 2028, driven by demand for chips used in low-Earth orbit satellite networks.
CEOs from ASML, Siemens, and five other tech companies banded together in an opinion piece arguing Europe must simplify its digital rules, scale homegrown technologies such as semiconductors and AI, and act faster to protect its global competitiveness and technological sovereignty.
Global semiconductor sales reached ~$299B in Q1 2026, up 25% from Q4 2025, reportsSIA. March worldwide sales were ~$100B, increasing 79% year-over-year and ~12% sequentially, with growth across all major regions.
Research
An MIT team created a silicon-photonics lidar chip design that significantly lowers signal crosstalk, enabling wider field of view, compact solid-state lidar sensors for AVs, drones, and other sensing systems.
UC Berkeley-led researchers transformed titanium dioxide (TiO₂) into a ferroelectric material by reducing its thickness to less than 3nm, opening a path to ultra-scaled, energy-efficient electronic devices.
Think tank Interface argues that Europe’s AI push could be limited by power-grid bottlenecks and inefficient accelerator use, so AI data centers should be planned around energy availability, grid flexibility, and compute-per-watt—not just chip supply.
The EU and Japan agreed to deepen digital cooperation across AI, data, quantum, semiconductors, digital infrastructure, and online platforms during the fourth EU-Japan Digital Partnership Council meeting in Brussels.
Asia
India’s government approved two more semiconductor projects, including Crystal Matrix’s GaN-based mini/micro-LED display and compound semiconductor facility in Dholera, Gujarat, and Suchi Semicon’s OSAT facility for discrete semiconductors in Surat.
RadixArk launched with $100M seed to build open infrastructure for Frontier AI.
Qutwo raised €25M in an angel round to build a European AI lab for the quantum era.
More partnerships
Nvidia and Corning announced a multi-year partnership to expand US-based manufacturing of optical connectivity for AI infrastructure.
Sony and TSMC will collaborate on the development and manufacturing of next-gen image sensors.
Xanadu and EV Group announced a deal to develop heterogeneous integration and wafer bonding processes for industrial-scale photonic quantum HW.
Fraunhofer‘s FhCAP photonics division is teaming up with Aegiq to scale photonic quantum HW.
Anthropic will use all of SpaceX’s compute capacity at its Colossus 1 data center, providing access to ~300 megawatts of new capacity — enough to power 220,000 Nvidia GPUs.
Reports
Mature-node foundry pricing is beginning to rebound as AI servers, general-purpose servers, and edge AI drive demand for PMICs, power discretes, and other power-related processes, reports TrendForce. The firm expects average 8-inch capacity utilization among the top 10 foundries to approach 90% in 2026 and remain above 80% through the first half of 2027, while 12-inch mature-node capacity adjustments could create order spillover and pricing opportunities for second-tier foundries.
Rambus introduced PCIe 7 switch IP with time division multiplexing, enabling efficient, scalable PCIe (Peripheral Component Interconnect Express) fabrics for data center infrastructure, distributed AI clusters, and HPC networks.
Toshiba launched 2:1 multiplexer and 1:2 demultiplexer switches to support high-speed interfaces such as PCIe 6 and USB4 V2.
Fig. 1: Rambus’ PCIe 7.0 switch IP with time division multiplexing. Source: Rambus
Uviquity rolled out a chip-scale deep-UV laser operating at 229nm, built on its aluminum nitride PIC platform.
Siemens EDA’s Veloce Strato CS hardware-assisted verification platform was used to verify Arm’s AGI CPU from subsystem through full-system level, supporting validation of key performance metrics for hyperscale deployments.
SiTime launched its Elite 2 Super-TCXO oscillator to boost GPU utilization and compute efficiency in AI data centers by delivering better time synchronization.
Research
A USC team created a framework for designing RFICs, using “a class of algorithms called metaheuristics, which explore enormous design spaces through guided randomness without needing prior examples.”
Researchers at Argonne and Northwestern U. collaborated on designing and predicting new materials by controlling and understanding how the atomic structures form.
AI is being integrated across the semiconductor ecosystem at a pace that far outstrips any rules to govern it, raising the specter of increased IP theft and security breaches with no obvious way to prevent them.
CISPA warns that on-body devices like smartwatches, smart glasses, VR headsets, and smart textiles pose major privacy and safety risks because they collect excessive amounts of high-resolution data that enable inferences about the wearer.
Numerous universities, including Rutgers, Baylor, and Duke, have been hit by a major ransomware attack on the Canvas platform, owned by Instructure.
Government intervention
NIST‘s Center for AI Standards will conduct pre-deployment evaluations on AI frontier models from Google DeepMind, Microsoft and xAI, aiming to understand capabilities and protect national security before they are publicly available.
CISA unveiled a new initiative called CI Fortify, which aims to help guide critical infrastructure entities across all sectors to prepare for crisis or conflict. There were also additions to the Known Exploited Vulnerability catalog.
AAA’s recent study found that extreme climates hurt both EV and hybrid efficiency, with cold weather having the biggest impact. At 20°F, EVs lost 35.6% in MPGe (miles per gallon of gasoline equivalent) and 39.0% of calculated range, compared to hybrids losing 22.8% in fuel economy. At 95°F temperatures, EV efficiency fell 10.4%, and hybrid MPG fell 12.0%.
Keysight and CATARC will collaborate on a Joint Innovation Laboratory for Charging Test Technology to address charging compatibility, safety, overseas compliance, and charging infrastructure.
Quantinuum and BMW have expanded their collaboration to advance future mobility by applying quantum computing toward advanced materials science for the design and optimization of fuel cells.
Battery management systems are growing increasingly smarter with innovations in software and hardware that enable more accurate estimation of battery state of charge and health, along with predictive diagnostics.
Ford provided a glimpse into its new EV Development Center in California, with a $30,000 or so Universal EV truck slated for 2027.
Nir Sever at proteanTecs talks about the complexities of moving data between different types of chiplets, what kinds of physical effects engineering teams need to consider when placing chiplets next to other chiplets, and how to balance performance and power across chiplets before they are packaged together to ensure the whole package remains reliable throughout its expected lifetime.
Workforce, Education
Stony Brook University launched its NSF-supported program to train students in semiconductor and advanced nanomanufacturing skills, using paid hands-on experience, a Brookhaven National Lab collaboration, and industry partners including GlobalFoundries and Wolfspeed to help fill the U.S. semiconductor workforce gap.
Over 120 students participated in the University of Sheffield’s “UK Tiny Tapeout,” submitting their own chips for fabrication.
InP and SiPho join CMOS as critical technologies. Lasers, CPO and OCS will be everywhere (indium phosphide, silicon photonics, co-packaged optics, optical circuit switch).
Reducing variation in manufacturing, monitoring behavior over time, and targeting specific workloads can have a big impact on power, performance, and area/cost.
This website uses cookies to improve your experience while you navigate through the website. The cookies that are categorized as necessary are stored on your browser as they are essential for the working of basic functionalities of the website. We also use third-party cookies that help us analyze and understand how you use this website. We do not sell any personal information.
By continuing to use our website, you consent to our Privacy Policy. If you access other websites using the links provided, please be aware they may have their own privacy policies, and we do not accept any responsibility or liability for these policies or for any personal data which may be collected through these sites. Please check these policies before you submit any personal information to these sites.
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
Any cookies that may not be particularly necessary for the website to function and is used specifically to collect user personal data via analytics, ads, other embedded contents are termed as non-necessary cookies. It is mandatory to procure user consent prior to running these cookies on your website.
Leave a Reply