Today's Semiconductors Pulse
SMIC clears final hurdle for $5.97B takeover of SMNC
China’s Semiconductor Manufacturing International Corp (SMIC) secured regulatory approval to acquire the remaining 49 % of its Beijing‑based unit SMNC, valuing the stake at 40.6 billion yuan (≈$5.97 bn). The transaction will be funded by issuing 547.2 million A‑shares at ¥74.20 ($10.91) each, with a 12‑month lock‑up for sellers, making SMNC a wholly‑owned subsidiary.
Tohoku University AI System Automates Quantum‑Dot Voltage Tuning, Boosting Qubit Scaling
A team from Tohoku University has demonstrated an AI‑driven workflow that automatically extracts charge transition lines from quantum‑dot measurements, removing the need for labor‑intensive manual tuning. The breakthrough, detailed in Scientific Reports, could accelerate the path to large‑scale semiconductor qubit arrays.
Protests Threaten AI Compute Capacity, Cancel 40% of Projects
Right now, because electricity prices keep going up, consumers have pinned this blame on AI Datacenters. This is unfair but there is no organized or concerted effort being put in to turn this tide. As the perception of AI and...
Samsung Electronics Produces First Working Die on Sub-10nm DRAM Node
Samsung Electronics announced the first working die on its sub‑10 nm DRAM process, dubbed the 10a node. The die uses a 4F² cell layout and a vertical channel transistor (VCT) structure, shrinking the cell footprint to roughly 9.5‑9.7 nm. Samsung aims to...
Report: Samsung Execs Worried Company Could Lose Money on Smartphones for the First Time
Samsung's mobile division warned it could post its first net loss on smartphones in 2026 despite strong Galaxy S26 sales. The loss is driven by soaring DRAM and NAND prices, which now account for a larger share of component costs,...
Why the Party for Intel and Other Chip Stocks Could Last a Long Time
Intel’s stock has surged 81% year‑to‑date and jumped another 25% after a surprise first‑quarter earnings beat that lifted gross‑margin and server‑chip revenue. The results pushed Intel’s forward price‑to‑sales ratio down to 7.1, making it cheaper than peers like Nvidia, Broadcom...
ARKothari Explores Texas Plant to Supply High‑Purity Semiconductor Chemicals
ARKothari Corporation announced a feasibility study for a high‑purity semiconductor chemicals plant in Texas, targeting domestic chip makers. The effort, backed by potential partners such as Exyte, Jacobs and Air Liquide, underscores a push for supply‑chain localization amid ongoing chip...
MaxLinear Q1 Revenue Jumps 43% as Enterprise Data Center Chip Demand Surges
MaxLinear reported first‑quarter 2026 revenue of $137.2 million, up 43% year‑over‑year, as infrastructure chip sales for enterprise networking surged 136%. The company lifted its 2026 optical data‑center revenue outlook by $30‑$40 million, citing strong visibility into hyperscaler program ramps.
Intel Shares Surge Nearly 20% After Blockbuster Q1 Earnings Beat
Intel reported Q1 revenue of $13.58 billion and EPS of $0.29, beating forecasts and propelling its stock up 19.9% to $80.10. The earnings beat fuels optimism across large‑cap stocks as the S&P 500 nears record highs.
US Futures Slip on Iran Talks, Boosting Pressure on Asian Markets as SK Hynix Posts Record Profit
US equity futures slipped 0.1% as Iran‑U.S. negotiations stalled, reigniting concerns over Middle‑East volatility. The dip hit Asian markets hard, even as South Korea’s SK Hynix posted record quarterly profits, underscoring a split between macro risk and strong corporate earnings...
AI Slashes Nvidia Chip Design From Months to Nights
Nvidia’s Chief Scienst Bill Dally at GTC explained how Nvidia uses AI to speed up the chip design process: ▫️trained an LLM on all proprietary internal Nvidia docs over past 30+ years (junior employees query it instead of interrupting senior designers) ▫️one...
The System Architect’s Sketchbook: The Buildout Frenzy
Deepak Shankar, founder of Mirabilis Design, released a cartoon titled “The system architect’s sketchbook: The buildout frenzy” on EDN. The illustration spotlights the accelerating pressure on electronics design engineers to integrate ever‑more functions, components, and software into a single system....
NEO Semiconductor's Revolutionary 3D X-DRAM for AI Processors Has Passed Proof-of-Concept Validation — Company Secures Funding to Develop Next-Gen Memory...
NEO Semiconductor announced that its 3D X‑DRAM technology cleared a proof‑of‑concept stage, delivering sub‑10 ns latency, over‑1 second data retention at 85 °C, and endurance beyond 10¹⁴ cycles using existing 3D NAND fab lines. The company also secured a strategic investment led by Acer...
Electronics Firm MELSS Forms JV with US-Based Valtech for Specialty Chemicals Used in Semiconductor Making
Indian engineering firm MEL Systems and Services (MELSS) has entered a joint venture with U.S. specialty‑chemical maker Valtech Corporation to produce chemicals and plastic components for semiconductor and solar‑cell manufacturing. The new entity, MELSS Valtech Ltd, will build a plant in Chennai...
Intel Boosts Margins by Selling Salvaged Edge-Die Chips
Got some clarity from Intel IR on additional lift to margins. Intel got an unexpected margin lift from better yield salvage. Chips that would normally have been lower-value edge-die on the wafer were binned down and still sold into usable...
Nextorage Unveils Fresh New Gaming SSDs, but They only Come with a One-Year Warranty
Nextorage has launched a new line of QLC gaming SSDs, offering capacities from 1 TB to 8 TB with write endurance ranging from 400 TBW to 4,000 TBW. The 1 TB model is priced at ¥44,550 (about $280), positioning it as a premium‑priced option in...
Nvidia’s Integrated Rubin Slashes AI Inference Costs Tenfold
5 Chip Stocks at the Center of the AI Spending Boom 👇 $NVDA Nvidia benefits from AI more directly than any other company because it sells the integrated system, not just the chip. Vera Rubin combines CPU, GPU, NVLink, ConnectX-9, BlueField-4...
CPU Requirements for AI Workloads Are Multiplying, Driving Intensifying Shortages and Price Hikes — Intel Already Shifting Production From Consumer...
Intel disclosed that AI workloads are driving a rapid shift from training to inference, tightening CPU‑to‑GPU ratios in data centers from roughly 1:8 to 1:4 and potentially to parity at 1:1. The surge in CPU demand has pushed server CPU...
Intel Shares Surge on AI Boom to Surpass Dotcom Bubble High
Intel’s stock jumped more than 12% in after‑hours trading, pushing the share price above $58 and eclipsing the peak it hit during the 2000 dot‑com bubble. The rally was sparked by strong earnings guidance and a surge in demand for...
Huawei, DeepSeek Strengthen China’s AI Self-Reliance with Collaboration on V4 Model
Huawei announced that its new Ascend 950PR and 950DT chips were ready for day‑zero inference on DeepSeek’s V4 model, with the entire Ascend SuperNode line fully adapted. The partnership was highlighted in a livestream where engineers detailed integration with Huawei’s...
DigiKey, STMicroelectronics, Ultra Librarian Provide Free, No-License Design Environment Online
DigiKey, STMicroelectronics and Ultra Librarian have launched eDesignSuite, a free, browser‑based design platform that hosts the world’s largest CAD library with millions of verified symbols, footprints and 3D models. The tool integrates thermal and electrical simulation, real‑time BOM impact analysis,...
Space Data-Center News: Roundup of Extraterrestrial AI Endeavors
In April 2026 a wave of space‑compute announcements pushed orbital data centers closer to commercial reality. Sophia Space teamed with Kepler Communications to run its software on a ten‑satellite cluster of Nvidia Orin processors, while Deloitte activated two additional cyber‑defense...
Aeluma Receives NASA Award for Integrated QD Lasers
Aeluma Inc., a Goleta‑based compound semiconductor firm, secured a NASA award to accelerate its integrated quantum‑dot (QD) laser platform for data‑communication and sensing. The company’s large‑diameter wafer heterogeneous integration leverages high‑throughput MOCVD to embed QD lasers directly on silicon, tackling...
Atomera Extends Collaboration with Synopsys to GaN Workflows
Atomera Inc. has broadened its long‑standing partnership with Synopsys Inc. to include gallium nitride (GaN) device modeling using Synopsys’ Sentaurus TCAD platform. The expanded effort will develop a GaN calibration methodology, produce marketing and calibrated simulation decks, and feed product...
AMD EXPO 1.2 Expands DDR5 Tuning Before Zen 6 Rollout
AMD unveiled EXPO 1.2, the latest iteration of its DDR5 overclocking profile standard. The update expands profile data, adds support for MRDIMM, CUDIMM and CSODIMM modules, and introduces new timing and voltage fields such as tREFI and VDDP. While early BIOS...
Intel Says Software Optimization Limits E-Core Gaming Performance Gains
Intel’s Vice President Robert Hallock says the gaming performance gap between Intel’s hybrid CPUs and AMD’s Ryzen X3D chips is largely a software issue, not a hardware flaw. Intel reports that toggling efficiency (E) cores changes game FPS by roughly one...
Intel Z970 Chipset Will Become Mainstream Segment for Nova Lake-S
Intel is set to launch the Z970 chipset alongside its Nova Lake‑S processors, positioning it as the core of the new 900‑series upper‑mainstream platform. The Z970 will inherit many Z‑series features—most notably CPU overclocking and richer I/O—traditionally reserved for enthusiast...
Middle-East Conflict Disrupts Helium Supply, Tightening AI Chip Production
The Iran war has crippled Qatar's helium production, a critical coolant for semiconductor manufacturing. With helium spot prices roughly doubling, memory‑chip giants Samsung, SK Hynix, Micron and foundry TSMC face longer lead times and margin pressure, raising alarms across the...
Google Launches TorchTPU, Letting PyTorch Run Natively on Massive TPU Clusters
Google announced TorchTPU, a software stack that enables PyTorch workloads to execute natively on its Tensor Processing Unit (TPU) hardware. The integration promises seamless migration with a single line change and supports the newly split TPU 8t and 8i families,...
SiTime Pushes Precision Timing Chips to Slash AI Data‑centre Power Use
SiTime unveiled precision timing silicon designed to tighten synchronization in AI clusters, promising measurable energy savings across thousands of servers. The move targets growing regulatory scrutiny of data‑centre power use and the climate‑impact of AI workloads.
VanEck SMH ETF Jumps 27% in 2026, Leaving S&P 500 Far Behind
VanEck's Semiconductor ETF (SMH) rose 27.73% year‑to‑date, climbing from $373.30 to $476.83, while the S&P 500 posted a modest 4.07% gain. The rally is driven by outsized gains in top holdings like TSMC, ASML and Intel, underscoring a broader thematic...
Tesla Admits Pre‑2023 HW3 Cars Can’t Deliver Unsupervised Full Self‑Driving, Offers Trade‑in Upgrade
Tesla announced that vehicles equipped with Hardware 3 (model years 2019‑2023) lack the memory bandwidth needed for unsupervised Full Self‑Driving. The company will offer a discounted trade‑in for HW4 (AI4) hardware and plans micro‑factories to retrofit cameras and computers, a...
Intel Shares Surge on AI‑Driven Sales Outlook, Reach 2000‑Era Peaks
Intel reported a 22% year‑over‑year jump in datacenter AI revenue and projected Q2 sales of $13.8‑$14.8 billion, sending the stock sharply higher and reviving confidence in its AI‑focused turnaround plan.
D-Mode GaN Bidirectional Switches Reshape Power Conversion Topologies
Renesas introduced the TP65B110HRU, a 650 V, 110 mΩ D‑mode GaN bidirectional switch that integrates two low‑voltage MOSFETs in a single TOLT package. The device conducts and blocks current in both directions, eliminating the need for back‑to‑back MOSFET or IGBT pairs. This...
Vishay Expands Power DFN Portfolio with Ultrafast Rectifiers
Vishay Intertechnology has added 16 new FRED Pt® ultrafast rectifiers to its Power DFN line, housed in the compact DFN6546A package and rated at 200 V with 6 A‑15 A currents. Both commercial‑grade and AEC‑Q101‑qualified automotive versions are available, offering a 10% lower...
Graviton Win Fuels $5B AWS CPU Revenue Growth
Solid win for Graviton, the most mature custom Arm CPU program of any hyperscaler. This business is around ~5B annualized revenue for AWS, and this will help escalate that number. AWS needs more CPUs. I know a foundry...
AMD Poised for Datacenter Dominance Ahead of Helios
AMD’s in such a great position. Exceptional datacenter CPUs, the IP and SOC architectures to do about any derivative a customer wants. And this is before Helios has kicked in. $AMD
AMD Ryzen™ Embedded V3000 Processor Powers Cisco’s Latest N9300 Switches and 8000 Series Routers for AI Workloads
Cisco has integrated AMD’s Ryzen Embedded V3000 processor into its new N9300 series switches and 8000 series service‑provider routers. The V3000 provides 4‑8 server‑class x86 cores, up to 96 GB DDR5 memory, and high performance‑per‑watt for control‑plane tasks such as routing, telemetry,...
First Look: Google’s Sunfish and Zebrafish TPUs
Our first pass deep dive on Google new Sunfish TPU 8t and Zebrafish TPU 8i. $GOOG https://t.co/7LELhFu6Oi
Intel's CPU Foundry Must Match TSMC to Meet Demand
I still maintain my CPU foundry of choice thesis here for $INTC. They don't have to be better than TSMC, just parity, and increasingly aid to fill CPU demand. https://t.co/dKqm16myKH
Be in the Know. 17 Key Reads for Friday…
Intel’s latest earnings beat expectations, sending the stock up 29% as AI‑driven demand for its CPUs accelerates. Analysts, including HSBC, raised price targets to $100, citing a resurgence in server‑grade chip sales. The AI rally spilled over to broader markets,...
Tenstorrent Accelerates Big MOE Models with V3/V4
This is cool. @tenstorrent is digging in. Big models are a passion project for us. May 1st we announce big numbers on V3, V4 results after that. Networked AI is a natural fit for big MOE models
AMD Likely to Use Intel Foundry Amid Capacity Crunch
I’m not going to die on this hill, for a ton of reasons lol but hard to see AMD not making CPUs at Intel Foundry before too long. Capacity is scarce, Intel is the floor space to ramp capacity.
The Overlooked Chips Powering the AI Boom
Power semiconductors, the unseen workhorses that convert high‑voltage electricity for AI chips, are poised for rapid growth as AI data centers consume ever more power. Omdia projects global revenue for power chips to reach $100 billion by 2029, up from $80 billion...
Customers Demanding Every Graviton Chip, Says Andy J
I suppose this was one of the customers who wanted ALL the Gravitons, per Andy Jassy.
US Government
The US government owns 9.8% of Intel and NVIDA owns another 4.5%. Not bad. https://t.co/UZRPnEpE5H
Chip Rally Lifts Qualcomm Stock Higher
The chip rally is so intense that even Qualcomm gets to go up $QCOM https://t.co/JHUQyG5MUs https://t.co/yQSvYsEvZ4
QCOM Poised for Lift with New DC CPU Launch
$QCOM should get a boost from launching a DC CPU. A much needed one. 💪🏻
Micron Set to Soar as Samsung Strike Spikes DRAM Demand
Few large specialists chiming in on $MU (memory) more bullish than ever. more server (or even client CPUs) means a lot more DRAM . MORE importantly SAMSUNG indicated a STRIKE, which cause $MU to surge https://t.co/7KhGWVDVdH
CPU Resurgence Fuels Gains for Intel, AMD, ARM, Qualcomm
$INTC, $AMD, $ARM, $QCOM all about to open higher on the great CPU resurgence. Expectations are $QCOM fleshes out their data center CPU/accelerator strategy soon.
Intel Emerges as the Go‑to CPU Foundry
$INTC because they are the only foundry who has line of sight to manufacturing expansion short-term. That, plus what I have been sharing for some time on my thesis Intel becomes the CPU foundry of choice for many. https://t.co/dKqm16myKH