Today's Semiconductors Pulse
Intel and Nvidia to launch integrated CPU‑GPU SoCs in early 2028
Intel and Nvidia have confirmed a joint effort to produce x86 system‑on‑chips that combine Intel CPU cores with Nvidia RTX GPU chiplets. The first products are slated for early 2028, potentially unveiled at CES. They will be fabricated on TSMC’s N3P process, feature LPDDR6 memory and use Nvidia’s upcoming Rubin architecture.
Tuurny Deploys AI Robotics to Harvest RAM Chips, Tackling US Chip Shortage
Tuurny unveiled its Nantul robotic cell, capable of pulling up to 300 RAM integrated circuits per hour from domestic e‑waste, and secured a six‑figure contract with UK processor Areera. The system promises a new on‑shore source of memory chips while curbing rare‑earth‑rich e‑waste exports.
Intel and Google Team up for Next‑gen AI Cloud Infrastructure
INTEL $INTC AND GOOGLE $GOOGL PARTNERSHIP Intel and Google just announced a multiyear collaboration to "advance the next generation of AI and cloud infrastructure"
HRL’s T3L 40nm GaN-on-SiC Technology Achieves Manufacturing Readiness Level 6
HRL Laboratories announced that its 40 nm T3L gallium‑nitride on silicon‑carbide (GaN‑on‑SiC) technology has attained Manufacturing Readiness Level 6, confirming production‑grade manufacturability under U.S. Department of War standards. The milestone paves the way for high‑volume manufacturing through a partnership with MACOM Technology...
Chip Can Project Video the Size of a Grain of Sand
Researchers from MIT, the University of Colorado, Sandia National Laboratories and MITRE have unveiled a sub‑0.1 mm² photonic chip that can steer light with unprecedented speed. The device uses voltage‑actuated metallic cantilevers to project up to 68.6 million light spots per second...
Advanced iBGA Image Sensor Encapsulation
DELO introduced a PFAS‑free encapsulation material designed for automotive image sensors used in advanced driver‑assistance and autonomous systems. The formulation delivers low warpage, fine‑pitch flow control, and automotive‑grade reliability, addressing the harsh temperature, vibration, and moisture conditions of vehicle environments....
Amazon CEO Says Chip Business 'On Fire' As AWS Steps Up Challenge To Nvidia
Amazon’s cloud unit reported a $15 billion annual AI revenue run rate and a custom‑chip business now exceeding $20 billion, double the Q4 figure. CEO Andy Jassy highlighted the rapid growth of its Trainium 3 accelerator, which promises better price‑performance than Nvidia GPUs....
Nvidia Rubin GPUs May Be Delayed, Slowing the Next Phase of AI Infrastructure
Nvidia’s next‑gen Rubin GPUs, slated for late‑2024 shipment, face supply delays due to HBM4 memory validation, CX9 interconnect rollout, and higher power and cooling demands. TrendForce now projects Rubin to represent only 22% of Nvidia’s 2026 shipments, down from an...
HighPoint Announces Rocket 1604L Compact PCIe Gen5 X16 Retimer AIC for AI and Industrial Edge
HighPoint Technologies unveiled the Rocket 1604L, a 4‑M.2 PCIe Gen5 x16 add‑in card powered by Astera Labs' Gen5 retimer. The compact 167 mm module delivers full 32 GT/s bandwidth with near‑zero latency, targeting AI edge and high‑velocity compute workloads. It integrates a Smart Firmware Layer...
Samsung to Invest $4 Billion in Chip Packaging Site in Vietnam
Samsung Electronics announced a $4 billion investment to build a chip‑packaging plant in Thai Nguyen province, northern Vietnam. The project will roll out in multiple phases, with an initial $2 billion tranche slated to start construction soon. The facility aims to boost Samsung’s...
GigaDevice Names ACP Distribution Partner for Brazil
GigaDevice, a leading semiconductor maker, has appointed ACP Componentes Eletrônicos as its strategic long‑term distribution partner for Brazil. ACP will handle the full GigaDevice portfolio—including flash memory, specialty DRAM, 32‑bit MCUs, sensors and analog products—across all 26 Brazilian states. The...
AI Semiconductor Market to Reach USD 1.1 Trillion by 2035
The AI semiconductor market is set to exceed $1.1 trillion by 2035, according to a new industry report. Valued at roughly $102.8 billion in 2025, it is expected to grow at a 27‑33% compound annual growth rate. Growth is fueled by expanding...
Samsung Advances 2nm GAA Push with Taylor Fab Targeting 2026 Start
Samsung Electronics is moving its 2nm gate‑all‑around (GAA) production to the Taylor, Texas fab, where engineers began equipment setup in March after receiving a temporary occupancy certificate. ASML has deployed EUV lithography tools to support the new node, a critical...
AMD Making It Easier To Embed Lemonade AI Capabilities Into Other Apps
AMD’s Lemonade open‑source AI server has released version 10.2, focusing on embeddable binaries for Linux and Windows. The new artifacts contain only the Lemond daemon, CLI and essential resources, removing web UI and Electron components. Lemonade 10.2 continues to support...
AI Will Fuel Bull Market Despite GPU Scarcity
"As the Iran War has ebbed and flowed, GPU availability for B200s has collapsed to zero...H100s close behind. Whatever happens with the war, the AI complex is likely to lead any true sustainable bull market" -Warren Pies, 3Fourteen Research
NDS Taiwan to Showcase Integrated Warpage Control Solution for PLP at SEMICON SEA 2026
NDS Taiwan unveiled an integrated warpage‑control solution for panel‑level packaging (PLP) at SEMICON Southeast Asia 2026. The approach coordinates stress‑compensating materials, temporary glass carriers, precision grinding, advanced polishing and clean debonding to manage deformation across the entire panel workflow. By...
Vector Photonics Demos Free-Space Optical Communication Using PCSEL Outside of a Lab
Vector Photonics showcased its photonic crystal surface‑emitting lasers (PCSEL) in a real‑world free‑space optical link across Glasgow’s River Clyde, transmitting 50 Mbps over 500 m. The trial, built with Fraunhofer UK, moved the technology from a lab‑only proof‑of‑concept to a commercial‑grade readiness...
Samsung Networks Boss Wonders if AI-RAN Is Too Hot to Handle
Samsung Networks CEO Woojune Kim warned that Nvidia‑style GPUs run so hot they could "cook a steak," raising doubts about their suitability for AI‑enabled radio access networks (AI‑RAN). He argued that RAN hardware must become cheaper and low‑power, a market...
US Forces Allies to Adopt Stricter China Chip Bans
So, after a war in Iran, threats to leave NATO, and zero focus on Asia, what position is Washington in to demand this of Japan or the Netherlands? Washington pushes allies to match tougher China chip curbs under new bill New US...
Early HBM4 Validation Points The Way For Next Generation AI And HPC Systems
Memory bandwidth is becoming the primary bottleneck for AI and high‑performance computing, driving the industry toward High‑Bandwidth Memory 4 (HBM4). Synopsys announced the world’s first HBM4 IP test chip that has been validated in silicon, achieving 9.2 Gbps eye‑opening performance across...
DRAM’s Whac‑A‑Mole Security Crisis
Rowhammer remains a pervasive DRAM security flaw, and a newer variant called Rowpress is emerging as a complementary threat. Memory manufacturers have introduced refresh‑management commands—RFM, ARFM and DRFM—to target vulnerable rows, yet these mitigations are imperfect and can be weaponized....
A New Era For Co-Processing
The semiconductor industry is shifting toward heterogeneous co‑processing architectures as AI workloads outpace single‑processor capabilities. CPUs remain the host, while GPUs, DSPs, NPUs and emerging RISC‑V accelerators handle specialized tasks, with data movement becoming the primary efficiency bottleneck. Vendors stress...
Fast Isn’t Fast Enough: Redefining Metrics for Edge AI
Industry leaders at Arm, Cadence, Rambus and others argue that edge AI performance is no longer measured by peak TOPS but by real‑world latency, power draw and memory efficiency. They note that data movement and bandwidth now limit inference more...
Redefining AI Inference With New Silicon Architecture
VSORA, a fabless semiconductor firm, unveiled its Jotunn8 and Tyr AI chip families built on a reimagined data‑movement architecture that dramatically lowers cost per query for hyperscale data‑center inference and powers demanding edge use cases such as autonomous driving. The...
Poland Enters the European Semiconductor Race. Important Agreement with France
Poland’s CEZAMAT research centre has signed a cooperation agreement with France’s state‑owned CEA‑Leti to develop fully depleted silicon‑on‑insulator (FD‑SOI) technology. The partnership joins the European Chips Act pilot‑line network alongside IMEC and Fraunhofer, giving Polish researchers access to world‑class expertise....
Intel and SambaNova Target Agentic AI Inference with Xeon 6
Intel and SambaNova announced a heterogeneous inference architecture that pairs GPUs, SambaNova’s Reconfigurable Dataflow Units (RDUs) and the upcoming Xeon 6 processor. The design allocates prefill, decode and control tasks to the hardware best suited for each stage, aiming to boost...
ThunderX Uses SAFERTOS for Cockpit-ADAS Fusion
ThunderX announced the integration of SAFERTOS, a pre‑certified real‑time operating system, into its next‑generation AI domain controller platforms built on Qualcomm SA8775P and SA8650P automotive SoCs. The safety‑focused RTOS will run in selected safety‑critical subsystems of the AIBOX family, supporting...
AMD Between the Memory Crisis and Price Stability: Is the 8GB Radeon RX 9060 XT a Damage Control Measure or...
AMD officially launched the Radeon RX 9060 XT on May 20 2025 with two memory options – an 8 GB model priced at $299 and a 16 GB version at $349 – targeting the 1440p segment. At the same time, a global DRAM price surge of roughly...
Fujitsu at the Core of Japan’s AI Independence Drive
Fujitsu is developing an AI inference device that merges its own neural processing units (NPUs) with Arm‑based Monaka CPUs, slated for production at Rapidus using a 1.4‑nm process co‑developed with IBM. The chip is designed for ultra‑low power consumption—100‑1000× less...
Framework Warns of Further Increases in RAM and SSD Costs: The Memory Crisis Is Affecting Repairable PCs as Well
Framework updated its April 6 price list, openly warning that DRAM, LPDDR5x and NAND costs are only temporarily stabilised and will likely rise through the rest of 2026. The company left DDR5 RAM prices unchanged but raised the 4‑TB WD Black...
AMD Ryzen 9 9950X3D2 Listed in Stores Ahead of Launch: Early Store Listings Show a Release Window, but No Confirmed...
AMD’s new Ryzen 9 9950X3D2 dual‑edition CPU is appearing in retailer catalogs ahead of its official launch, with multiple stores listing an April 22 pre‑order window. The processor retains the 16‑core/32‑thread Zen 5 core count but expands L3 cache to 192 MB and raises TDP...
SEMI: Global Semiconductor Equipment Billings Up 15% YoY in 2025
SEMI reports global semiconductor equipment billings rose 15% year‑over‑year to $135.1 billion in 2025, up from $117.1 billion in 2024. Front‑end wafer processing equipment grew 12% while other front‑end segments rose 13%, and back‑end test equipment surged 55% as AI and high‑bandwidth...
AI Turns Power Grid Into Semiconductor Bottleneck
#Technology #Thread #Semiconductor #Manufacturing #AI The Semiconductor AI Infrastructure Grid: 1/ - Artificial Intelligence Is Not Just Powering Chips It Is Reshaping The Grid That Powers Semiconductor Fabs And Data Centers. - Rising AI Workloads Plus Advanced Semiconductor Fabs Equal Surging Electricity Demand. -...
Amazon and Nvidia Share Israeli AI Chip Roots
For those of you blown away by $AMZN announcement on Trainium4 and $NVDA growth, they have something important in common…🇮🇱🇮🇱
AMD Announces Ryzen 9 9950X3D2 Pricing and Availability
AMD unveiled the Ryzen 9 9950X3D2 Dual Edition, a desktop processor featuring dual 3D V‑Cache. The chip launches on April 22 with a suggested retail price of $899. Targeted at workstation users, it aims to deliver higher cache density for compute‑intensive workloads. AMD...
Intel Missed Chance by Denying Gelsinger's Foundry Investment
Can you imagine if Intel's board let @PGelsinger invest in the foundry capacity he asked them for? I agree, it seemed crazy at the time, but hindsight is 20/20.
TSMC and Top Memory Makers Set Prices, Customers Agree
TSMC, and the three memory names right now, are the primary beneficiaries of a name your price market dynamic. Meaning they can literally name their price and customers will pay it.
Foundryecosystem Report: Nvidia GPU Delays; Tools; IC Prices
The latest Foundryecosystem Report highlights several critical shifts in the semiconductor sector. Nvidia’s new Rubin GPU ramp is trimmed to 1.5 million units after HBM4 qualification setbacks at SK Hynix and low yields at Micron, delaying mass production to September. Applied Materials...
Amazon Nova 2 Still Lags Behind Sonnet 4.5
So what's the deal with Amazon Nova? They released Nova 2 in December, and even then, the top flight Nova 2 model trailed Sonnet 4.5. And it still hasn't left preview.
SolarWindow Technologies Inc (WNDW) Q2 2026 Earnings Call Transcript
Wolfspeed reported Q2 FY2026 revenue of $168 million, aligning with the midpoint of its prior guidance, while AI data‑center sales surged 50% quarter‑over‑quarter. The company ended the quarter with $1.3 billion in cash after a $700 million tax‑credit refund, but posted a...
Google Cloud Axion for Snowflake Gen2 Warehouses: Next-Generation Price Performance and Memory Bandwidth
Google Cloud and Snowflake have integrated Google’s new Axion processors—delivered via C4A virtual machines—into Snowflake Generation 2 warehouses. The custom Arm‑based silicon and DDR5 memory promise up to 50% faster query performance and a similar boost in memory bandwidth. Early adopters...
Nvidia Vs. Micron: Which AI Chip Stock Has More Upside Potential?
AI’s rapid expansion has heightened demand for both GPUs and high‑bandwidth memory, creating a tightly linked market for Nvidia and Micron. Nvidia’s upcoming Vera Rubin platform promises up to 75% fewer GPUs for training and a 90% reduction in inference...
Chinese Memory Makers Gain Share with Cheaper, Higher‑capacity Chips
Chinese memory giants seen grabbing market share on lower prices, more capacity Leading Chinese memory producers Yangtze Memory Technologies Co (YMTC) and ChangXin Memory Technologies (CXMT) are working on boosting capacity this year. https://t.co/erh2vmYWgO
Alibaba Deploys 10,000-Node AI Cluster with Zhenwu Chips
As AI race with US intensifies, Alibaba launches 10,000-card computing cluster powered by the Zhenwu AI chips developed by its T-Head semiconductor design arm. https://t.co/XHmAX5hhjW
Momentum-Engineered Photonic States Make Bulk Silicon Shine
Researchers at UC Irvine have shown that bulk silicon can emit bright, broadband light by engineering the momentum of photons rather than altering the material itself. By decorating silicon surfaces with sub‑2 nm metal particles, they create extreme light confinement that...
US Chip Curbs Threaten Supply Chain, Timeline Unrealistic
Proposed US chip curbs on China risk deeper supply chain disruption ...whether the proposed 150-day timeline was sufficient to achieve meaningful policy coordination remained uncertain.... Clearly not feasible. It has been nearly 4 years already... https://t.co/RLwyFRVVUe
RMBS Spotlight: Labs’ IP Accelerates Memory Industry
I heard people may be starting to look into $RMBS. I published (in February) an interview with their SR Fellow of Labs on how the memory industry is being pushed forward with their IP. Hope it helps. https://t.co/nrz9hxwGHR
Contemplating Meta’s Homegrown MTIA Compute Engine Roadmap
Meta Platforms is advancing its home‑grown MTIA AI compute engine to support next‑generation generative recommenders built on the Hierarchical Sequential Transduction Unit (HSTU). The roadmap introduces MTIA 100/200 (GPU‑like monolithic dies) and higher‑end MTIA 300‑500 models that use multi‑chip designs, HBM3/4 memory,...
Researching for DiligenceStack Pushes Me Into New Tech
I like how writing @DiligenceStack forces me to go do research I wouldn't normally have to do on the client side. But it's making me go from from the weeds of semis and now to the soil. All of...
Meta's MTIA AI Compute Engine Roadmap Unveiled
I finally had time to circle back to the Meta MTIA AI compute engine roadmap. $META $NVDA $GOOG $AMD https://t.co/0K4g5DFB3P
How Dummy Silicon Helps Drive 3D Chip Innovation
The semiconductor industry relies on purpose‑built test vehicles and daisy‑chain testing to validate 3D IC manufacturing processes. Test vehicles act as dummy silicon platforms that replicate critical interconnects, allowing engineers to probe thermal, mechanical and electrical reliability before full‑scale production....