Magnachip Expands Gen6 Super-Junction MOSFET Line with 15 Devices

semiconductor-digest.com Oct 6, 2026

Magnachip Expands Gen6 Super-Junction MOSFET Line with 15 Devices

Curated by Shubham · TechCapsules · Oct 6, 2026

AI-summarised brief · reviewed before publication

Magnachip Semiconductor expanded its power‑semiconductor lineup with 15 new sixth‑generation (Gen6) E6 super‑junction MOSFETs, targeting consumer and computing power applications. The devices, available in 500 V and 600 V ratings and multiple packages (DPAK, TO‑220FA, TO‑220SFA), balance thermal, switching, and EMI performance to meet the demands of compact, high‑efficiency systems. A 600 V, 280 mΩ E6 MOSFET demonstrated a 5.3 % temperature reduction in a TV power‑factor‑correction stage compared to the prior generation, underscoring its improved thermal margin. The series is now in mass production for TV SMPS, notebook adapters, PC power supplies, and sub‑1 kW industrial applications.

💡 Why It Matters

· The E6 Series equips designers with tighter thermal and noise control, enabling smaller, more efficient power modules that can meet rising performance targets in consumer electronics.

Capsules Semiconductors

Read full article on semiconductor-digest.com

Etched fields funding offers at $40B+ valuation, sources say

Etched, an AI chip startup, has attracted offers ranging from $40 billion to $50 billion, nearly double its recent $21 billion valuation after a $700 million raise. The company is pursuing costly full AI hardware systems powered by proprietary chips, with early funding talks still in progress. Investors are drawn by Etched’s potential to challenge Nvidia, its customer base including Jane Street, and its in‑house chip designs that promise faster, cheaper inference. A successful deal could extend the firm’s runway by up to 3.5 years.

💡 The surge in bids signals a shift in the AI hardware race, positioning Etched as a serious contender to Nvidia’s dominance. This could accelerate the deployment of more efficient, cost‑effective AI systems across finance and other high‑speed sectors.

androidcentral.com · Oct 6

Qualcomm and Arm are back in court, and your next phone is…

Qualcomm and Arm have resumed litigation in Delaware, launching a five‑day jury trial that could reshape smartphone chip licensing. Qualcomm seeks to suspend royalty payments for up to five years, citing Arm’s refusal to provide Nuvia testing tools and a leaked threat to terminate a key license. The dispute centers on Qualcomm’s 2021 Nuvia acquisition and a $170 million reduction in a Meta chip deal. Arm argues the changes stem from Meta’s pivot to AI glasses, not Qualcomm’s actions. The trial may alter future CPU technology negotiations and royalty structures.

💡 The outcome could redefine how major chipmakers negotiate licensing terms, potentially shifting control over next‑generation processor designs.

semiconductor-digest.com · Oct 5

Rapidus Builds Design Ecosystem Around Planned 2nm Foundry

Rapidus Corporation announced the creation of the Rapidus Collaborative Open Rapid Ecosystem (CORE), a global framework aimed at accelerating the development and mass production of 2nm semiconductors. The initiative partners 17 design firms—including Cadence, Synopsys, and Wipro—to provide design resources, expertise, and early-stage collaboration for customers worldwide. CORE will expand to include EDA tools, IP, and other ecosystem components, positioning Rapidus to support the entire semiconductor value chain from research to manufacturing.

💡 By uniting design, tooling, and manufacturing partners, Rapidus addresses the talent and integration gaps that have slowed 2nm progress, potentially reshaping the global semiconductor supply chain.

semiconductor-digest.com · Oct 5

JEOL Launches Low-Angle Ion-Milling SEM for Advanced-Chip Failure Analysis

JEOL Ltd. announced that its new low‑angle ion‑milling scanning electron microscope, named HAXIS, will go on sale on October 5, 2026. The instrument merges JEON’s passive voltage contrast (PVC) detection with a proprietary low‑angle argon ion‑milling delayering process to produce a flat, minimally damaged sample surface for semiconductor failure analysis. By preserving surface integrity, HAXIS delivers clear, reproducible PVC images that locate electrical anomalies in advanced chips without the probing artefacts that can alter measurements. The system’s integrated workflow—combining LIM delayering, SEM imaging and PVC analysis—aims to shorten the turnaround from defect detection to root‑cause identification, supporting QTAT analysis. JEOL cites reduced running costs from the argon gas and improved sample preparation as key advantages for the increasingly complex, nanoscale device market.

💡 Faster, more accurate defect localization lets chip makers cut costly debug cycles, accelerating time‑to‑market for next‑generation processors.

teslarati.com · Oct 5

Elon Musk teases TSMC as potential Terafab partner

Elon Musk confirmed that early discussions are under way with Taiwan Semiconductor Manufacturing Company (TSMC) about joining his Terafab semiconductor venture, a joint effort by Tesla and SpaceX aimed at delivering over one terawatt of AI compute capacity per year. While no agreement has been reached, the talks could position TSMC as an anchor customer for future Texas fabs, contributing process expertise and capacity in exchange for capital and long‑term purchase commitments from Terafab. Intel, already on board since April 2026, will supply its 14A process, and existing suppliers Samsung and TSMC continue to meet near‑term chip needs for Tesla’s AI5 and AI6 products. The first phase of the project is budgeted at roughly $16.8 billion, with total costs expected to climb into the tens of billions across multiple stages.

💡 Securing TSMC’s expertise would give Terafab a credible path to scale AI‑chip supply beyond current industry limits, directly supporting Tesla’s upcoming robotics, autonomous‑vehicle, and space‑data initiatives.

newsable.asianetnews.com · Oct 5

Cerebras Stock Gets A Boost After OpenAI CEO Sam Altman Calls Chipmaker…

Cerebras Systems shares surged nearly 7% after OpenAI CEO Sam Altman publicly affirmed the company as a “close partner,” countering rumors that OpenAI would favor Nvidia GPUs for its GPT‑6.1 Ultrafast mode. Altman’s statement eased investor anxiety following a 20% decline and a 10% drop from the IPO price, compounded by the expiration of insider lock‑ups that added selling pressure. The chipmaker’s partnership with OpenAI remains central, with a multiyear agreement to deploy 750 MW of wafer‑scale systems and a recent Q2 revenue jump of 74% to $180.1 million.

💡 Altman’s endorsement signals confidence in Cerebras’ high‑speed inference technology, potentially stabilizing its valuation amid competitive pressures. The comment may also influence future AI infrastructure contracts, reinforcing Cerebras’ position in the rapidly evolving chip market.