Skip to content
Chimera readability score 72 out of 100, Expert reading level.

The memory maker rolled out a slew of announcements this week, including:
Raised its planned U.S. investment to more than $250B through 2035, an incremental $50B above what was announced last June, with an ultimate goal of producing 40% of its DRAM in the U.S.;
Planned new investments of $3B for U.S. IC supply-chain investments, including $500M in financing for GlobalWafers’ 300mm raw silicon wafer facility in Sherman, Texas, along with a 10-year wafer supply agreement;
Entered into a long-term strategic customer agreement with Ford, providing a supply of memory and storage solutions for the carmaker’s next-gen vehicle production.
Started construction on its $9.3B expansion of its Hiroshima Prefecture advanced memory chip facility.
Large Deals
Apple announced a $30B+ multiyear commitment with Broadcom to design and produce custom silicon components and wireless connectivity technologies for a wide range of Apple products. To supply RF and wireless components, Broadcom will undertake a $1.5B expansion of its Fort Collins, Colorado, manufacturing facilities.
Solstice Advanced Materials agreed to acquire Element Solutions for $14.5B, adding Element’s electronics chemicals portfolio to Solstice’s specialty-materials business.
SEALSQ and GlobalFoundries will co-develop secure semiconductor platforms, focusing on PQC, cryogenic CMOS technology, and trusted chip production.
Onsemiplans to shed its Tarlac, Philippines manufacturing site to Greatek Electronics and its Mountain Top, Pennsylvania site to Silex Microsystems.
QuantumDiamonds received US$104M to ramp up production of its quantum-based semiconductor inspection technology.
SambaNova raised $1B to expand capacity for its on-prem agentic inference chips that use a dataflow architecture.
Oratomic raised $300M to build a fault-tolerant quantum computer based on reconfigurable atomic arrays trapped in focused laser beams.
New analysis
SIA/WSTS reports global semiconductor sales hit a record ~$121B in May 2026, up over 9% from April and over 104% from May 2025.
A NNME/SEMI/NSF/McKinsey analysis found the U.S. semiconductor industry will need 189K additional workers by 2030, with 74% of the projected demand related to manufacturing. The report noted that 75% of surveyed employers had difficulty hiring engineers.
TetraMem and SK hynix demonstrated progress on analog in-memory computing for AI, using resistive memory to process data closer to where it is stored.
Flex and Cerebras expanded U.S. production of Cerebras’ AI supercomputers, supporting an expected 7X production ramp for wafer-scale AI systems through 2026. Cerebras also said it will bring its first European AI data-center capacity online by the end of this year.
The Economist tested 25 leading AI models on morals and beliefs, finding that many don’t reflect the values held by most people around the world. That gap matters most when AI is used for judgment calls — in areas like advice, policy, education, or hiring — where a model’s built-in assumptions can shape the answer without users realizing it.
Legal matters
The U.S. ITC reaffirmed a previous patent finding that Innoscienceinfringed on an Infineon patent concerning GaN technology, resulting in import and sales bans against the Chinese semiconductor developer.
Wolfspeed filed a patent-infringement lawsuit in Delaware, alleging Navitas’ GaN-based FETs and other products infringe on multiple Wolfspeed patents. Navitas disputes the allegations.
Around the Valley
Amkor SVP David McCann (L) and ASE VP of R&D CP Hung (R) discuss future challenges and directions for advanced packaging at iMAPS CHIPCon conference in Santa Clara, California.
Financials
Samsung (Q2 guidance), TSMC (June sales) and UMC (June revenue).
The EU’s Chips JU opened its latest round of applications for projects that aim to advance research, infrastructure, power electronics, photonics, health, 6G, and international cooperation.
Harvard’s Belfer Center said the EU’s proposed Tech Sovereignty Package — including Chips Act 2.0 and the Cloud/AI Development Act — marks a move from only subsidizing European fabs toward building demand for advanced chips across strategic sectors. The brief says the effort is more likely to strengthen niche AI-chip design and ecosystem capabilities than to dramatically raise Europe’s share of chip manufacturing.
POLYN Technology and Alter Technology plan to establish a joint chip-validation lab in Toulouse, France, with specialized electrical validation capabilities for neuromorphic analog signal-processing chips.
Accellera released a draft version of the upcoming Functional Safety Language standard for public review through August 14. FSL defines the syntax and semantics of a language and data model for representing and exchanging functional safety intent.
Rambus introduced a DDR5 server RDIMM chipset operating at up to 9600 MT/s for AI infrastructure and HPC.
Samsung Electronics started mass production of a 4-16TB PCIe 6.0 SSD for AI and HPC server environments.
Infineon will supply 1,200 V CoolSiC MOSFETs and matching dual-channel EiceDRIVER 2EDB9259Y gate drivers to Advantics for its new line of liquid-cooled power converters.
Keysight updated its software test tool to find UI elements by description rather than visual appearance.
Daikin and NTT DATA launched a proof of concept for AI-based data-center cooling optimization, combining server thermal prediction with integrated control of HVAC, chillers, and liquid cooling for AI server environments.
Research
MITRE, UCBoulder, and other researchers built piezo-optomechanical photonic circuits directly on commercial CMOS driver wafers, creating a monolithic platform for dense electronic control of silicon nitride photonics used in AI, sensing, and quantum-computing systems.
Keysight announced the APS-ONE-400, a new modular network cybersecurity test platform. This new 4x100GE platform increases performance for Layer 4-7 traffic, encrypted traffic, and Elephant Flow traffic, all within a 1-rack unit system.
Government on cybersecurity
The U.S. Department of Defense opened applications for its new cyber apprenticeship program after receiving more than 70,000 applicant inquiries since the program was first teased in April.
The European Commissionpublished a new plan for AI security, setting new standards for AI evaluation, research, and cyber defense best practices.
The UK government published a frontier-AI defense plan called Cyber Shield, calling on organizations to ensure rapid patching of vulnerabilities, reduce reliance on legacy systems, and adopt secure-by-design technologies.
Warnings, vulnerabilities
CISA published a post-mortem on lessons from its own cyber incident, warning organizations should restrict who can upload code or files to public repositories, monitor repositories for exposed secrets, and ensure exposed credentials can be quickly detected, revoked, and replaced. The cyber agency also posted new vulnerability alerts this week.
Cloudflare said post-quantum signature migration cannot wait for better algorithms, arguing that ML-DSA’s size and implementation tradeoffs are acceptable for the first transition because newer NIST candidates are unlikely to be widely available before the 2030s.
Cisco and Foundry predict AI will triple network traffic over the next three years, driving growing concerns over network infrastructure and security risks.
Nearly a year after federal EV credits were terminated, first-time EV buyers in California can now get an instant $3,500 rebate for new cars under $50k.
Fiat introduced a tiny all-electric Topolino to the U.S. market, pricing the two-seat neighborhood EV starting at $13,995. However, it can only go 19-25 mph, has a maximum range of just 46 miles, and functions more like a golf cart.
Just because it’s a standard doesn’t mean everything will work together, and this is especially evident with conflicting USB standards. Cadence’s David Shin explains where incompatibilities can crop up, why it’s so difficult to control the different versions, and what’s behind all this confusion.
Quantum
Quantum technology is at a stage where maximum creativity meets maximum chaos. The industry hasn’t settled on a common approach to quantum computing, so different research groups have been working on different implementations simultaneously. Numerous small startups have sprung up, each addressing some aspect of a quantum system.
IQM Quantum Computers was listed on Nasdaq following the completion of a SPAC deal. IQM also acquiredQuantistry, a provider of AI-powered quantum chemistry and materials simulation software.
QpiAI open-sourced its SDK, which provides a Python-based interface to build and run quantum algorithms.
Quantum research
ETH Zurich researchers designed quantum memory using mechanical resonators that vibrate when storing information.
Oak Ridge National Laboratory, Cleveland Clinic, and IBM used a quantum computer to calculate nine molecular configurations of a potential fusion energy material.
Government on quantum
The U.S. NSF launched an initiative to integrate quantum sensing, quantum networking, and quantum computing into a single operational system.
Illinois will offer $3M to encourage NSF X-Labs quantum research teams to set up in the state.
Purdue expanded its semiconductor professional education portfolio with Taiwan’s NYCU, adding joint stackable industry courses in AI chip design, packaging, reliability, characterization, and more.
SEMI warns that skilled trades are becoming a big roadblock on the IC construction buildout, with large fab projects requiring thousands of workers and more specialized training in cleanroom construction, electrical systems, safety, and contamination control.
Florida’s St. Petersburg College received $1M in federal funding for its SMART Tech Industry 4.0 Lab, which will support semiconductor, mechatronics, automation, and robotics technician training.
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

Facts Only

* Memory maker raised planned U.S. investment to over $250B through 2035.
* Goal is to produce 40% of DRAM in the U.S.
* Planned $3B for U.S. IC supply-chain investments.
* $500M financing for GlobalWafers’ 300mm raw silicon wafer facility in Sherman, Texas, including a 10-year wafer supply agreement.
* Entered a long-term strategic customer agreement with Ford for memory and storage solutions.
* Started construction on a $9.3B expansion of the Hiroshima Prefecture advanced memory chip facility.
* Apple committed $30B+ to Broadcom for custom silicon components and wireless connectivity.
* Broadcom will expand manufacturing facilities by $1.5B in Fort Collins, Colorado.
* Solstice Advanced Materials agreed to acquire Element Solutions for $14.5B.
* SEALSQ and GlobalFoundries will co-develop secure semiconductor platforms focusing on PQC and cryogenic CMOS.
* SambaNova raised $1B to expand capacity for agentic inference chips.

Executive Summary

The memory maker increased its planned U.S. investment to over $250 billion through 2035, aiming to produce 40% of its DRAM in the U.S. It planned $3 billion in U.S. IC supply-chain investments, including a $500 million financing for GlobalWafers' 300mm silicon wafer facility in Texas with a ten-year wafer supply agreement. The company also secured a long-term strategic customer agreement with Ford for memory and storage solutions for next-generation vehicles and began construction on a $9.3 billion expansion of its Hiroshima facility. Other major developments include Apple's $30 billion commitment to Broadcom for custom silicon, and various strategic deals involving material acquisition, co-development in semiconductor platforms, and capacity expansions for agentic inference chips by SambaNova.

Full Take

The collection of announcements reveals a systemic push across the semiconductor ecosystem toward massive geopolitical realignment, vertical integration, and advanced computation architectures. The dramatic increase in investment by memory makers and the focused supply-chain financing indicate an effort to secure domestic and allied manufacturing capacity, particularly for cutting-edge technologies like DRAM and AI accelerators. Simultaneously, the interplay between fundamental research (quantum, analog in-memory computing) and immediate industrial application demonstrates a tension between long-term theoretical exploration and immediate commercial deployment. The persistent narrative surrounding talent shortages, as evidenced by workforce projections and industry concerns about skilled trades in fabrication, suggests that infrastructural scaling is bottlenecked not just by capital, but by the human and physical capacity to execute complex manufacturing processes. Furthermore, the philosophical gap noted in AI models regarding human values underscores a risk: technological advancement decoupled from broad societal consensus can lead to powerful systems operating on unaligned assumptions, raising questions about the governance needed for technologies like quantum computing and advanced AI judging. The integration of legal frameworks (patent disputes) alongside physical infrastructure builds a complex reality where market ambitions are constantly mediated by legal boundaries and existential security concerns regarding AI and cyber resilience.
Chip Industry Week In Review — Arc Codex