Today's Semiconductors Pulse
Intel to launch processor line with integrated NVIDIA GPU tech
Intel announced a new processor series that will embed NVIDIA GPU technology, aiming to combine CPU performance with RTX graphics. The integrated CPUs are slated for an early 2028 release, bringing high‑end GPU capabilities to standard Intel chips.
Snapdragon X2's Adreno X2-85 GPU Sees Driver Improvements For Linux 7.1
Rob Clark submitted a batch of MSM DRM driver changes for the upcoming Linux 7.1 merge window, targeting Qualcomm’s Snapdragon X2 laptop SoC. The updates bring preemption support, SKU detection with speed‑bin tables, and error fixes to the Adreno X2‑85 GPU, while the Adreno 840 gains preemption, SKU detection, and Inter‑Frame Power Collapse for power savings. Additional work adds Application Qrisc Engine (AQE) support for Vulkan ray‑pipeline capabilities and a suite of display‑related enhancements across Qualcomm GPUs.
Realtek RTL8159 10GbE to USB 3.2 Adapters Sell for About $55 and Up
Realtek's RTL8159 10 GbE‑to‑USB 3.2 adapters have entered the market at prices ranging from roughly $45 to $80, making multi‑gigabit Ethernet accessible for laptops and small‑form‑factor PCs. The WisdPi WP‑UT9 retails for $79 (about $87 shipped to the US), while the XikeStor...
Mobilint Raises 70 Billion Won in Series C Funding
Mobilint announced a 70 billion won Series C round on April 1, backed by Praxis Capital Partners, POSCO Investment and K Partners. The funding will fuel development of next‑generation neural processing units, expand mass‑production capacity, and support global market entry. Mobilint’s AI chip portfolio...
Qualitas Semiconductor Signs 2nm Edge AI IP License Deal Worth 1.3 Billion Won
Qualitas Semiconductor announced a design‑IP license agreement for its 2‑nanometer gate‑all‑around (GAA) MIPI C/D‑PHY, valued at roughly 1.3 billion won—about 21% of its projected 2024 revenue. The license, granted to a U.S. edge‑AI semiconductor firm, runs through March 30 2025 and carries no...
FuriosaAI to Mass Produce Second-Gen AI Chip, Supply to Samsung SDS Cloud From July
FuriosaAI announced mass production of its second‑generation Renegade AI chip, targeting 20,000 units in 2024 after an initial 4,000‑unit run. The chip upgrades to HBM3E memory, boosting capacity to 72 GB and delivering up to 20 petaflops per rack. Samsung SDS will...
MiTAC Shows Servers with Next-Gen CPUs and Solidigm SSDs at NVIDIA GTC 2026
MiTAC unveiled a next‑generation G‑Series server at NVIDIA GTC 2026, pairing dual AMD EPYC “Venice” CPUs with NVIDIA RTX Pro 6000 and 4500 Blackwell GPUs. The chassis places GPUs and Solidigm D7‑PS1010 PCIe Gen5 SSDs at the front, while NVIDIA ConnectX‑8 delivers eight 400 GbE...
C-Hawk Expands Southeast Asia Manufacturing Capabilities and Capacity with New Malaysia and Vietnam Facilities
C-Hawk Technology announced the opening of two new manufacturing facilities in Southeast Asia—a 200,000‑square‑foot plant in Johor Baru, Malaysia, and a 96,000‑square‑foot site in Ho Chi Minh City, Vietnam. The Malaysian location expands precision‑plastic production and adds PFA tube‑bending, while the Vietnamese factory...
IBM Announces Strategic Collaboration with Arm
IBM announced a strategic collaboration with Arm to create dual‑architecture hardware that blends IBM’s enterprise‑grade reliability and security with Arm’s power‑efficient designs. The partnership targets AI and data‑intensive workloads, emphasizing virtualization, high‑availability, and ecosystem growth. By integrating Arm‑based software environments...
GigaIO Sells SuperNODE and FabreX AI Fabric to D-Matrix
GigaIO announced the sale of its SuperNODE platform and patented FabreX PCIe Gen 5 AI fabric to low‑latency inference specialist d‑Matrix. The deal concludes a year‑long partnership that integrated d‑Matrix’s Corsair accelerators into SuperNODE, creating the industry’s most scalable inference node. Along...
Cohu Announces $30M Follow-On Orders for High Performance Computing Test
Cohu announced two customers placed follow‑on orders worth $30 million for its Eclipse platform with active thermal control, targeting next‑generation high‑performance computing (HPC) processors. The deals include a PAICe Prescriptive software subscription valued at roughly $330 K in annual fees, aimed at...
PrismML Launches World’s First 1-Bit AI Model to Redefine Intelligence at the Edge
PrismML emerged from stealth to launch the world’s first commercially viable 1‑bit large language model, the 8‑billion‑parameter Bonsai 8B. Built on Caltech research and trained on Google v4 TPUs, the model delivers performance comparable to full‑precision 8B models while using...
Embedded World 2026: Boards and Modules (Part 3)
Embedded World 2026 showcased five new edge‑AI boards and modules aimed at industrial and IoT markets. Variscite’s VAR‑SMARC‑MX8M‑PLUS brings a 2.3 TOPS NPU and dual Gigabit Ethernet, while congatec’s Conga‑SMX95 pairs an NXP i.MX 95 CPU with a 2 TOPS eIQ Neutron accelerator...
The Epoch Brief - March 2026
Epoch AI’s March 2026 brief highlights three new Data Insights, including a 4.1× annual rise in AI chip memory bandwidth now at 70 million TB/s, and reveals that advanced packaging and high‑bandwidth memory, not logic dies, constrained chip production in 2025. A...
AMD Details Upcoming Zen 6 PQOS Extensions: Advanced Bandwidth and Privilege Controls
AMD released a technical document outlining three new PQOS ISA extensions for its upcoming Zen 6 microarchitecture: Global Bandwidth Enforcement (GLBE), Global Slow Bandwidth Enforcement (GLSBE) and Privilege‑Level Zero Association (PLZA). GLBE lets system software set L3 cache external bandwidth caps...
Silicon Catalyst and Microelectronics US 2026
Silicon Catalyst has been named the exclusive strategic partner for the Microelectronics US 2026 conference in Austin, Texas. The two‑day event, scheduled for April 22‑23, will bring together more than 3,000 engineers, investors, and policymakers to discuss semiconductor design, AI hardware, advanced...
Mitsubishi Electric Enters ROHM–Toshiba Chip Integration Talks
Mitsubishi Electric has signed a memorandum of understanding to discuss integrating its power‑device operations with the semiconductor businesses of ROHM and Toshiba Electronic Devices, alongside Japan Industrial Partners and TBJ Holdings. The talks aim to create a globally competitive semiconductor...
This Chip Keeps Working at 700°C, Surviving Lava-Like Heat
USC researchers have built a memristor‑based memory chip that continues to function at 700 °C—hotter than molten lava—by stacking tungsten, hafnium oxide and a single‑atom graphene layer. The device stored data for over 50 hours without refresh, survived more than a billion...
NVIDIA Invests $2 Billion in Marvell to Expand NVLink AI Ecosystem
NVIDIA announced a $2 billion investment in Marvell Technology to broaden its NVLink ecosystem and support custom AI infrastructure. The partnership will deliver NVLink Fusion, a high‑speed interconnect that lets customers build rack‑level heterogeneous AI systems using NVIDIA GPUs, CPUs, DPUs...
Kioxia Killing Off Old NAND Chippery
Kioxia has issued an end‑of‑life notice for its third‑generation 3D NAND products, including 64‑layer planar and sub‑96‑layer chips such as SLC, MLC and TLC. Sales will cease in September 2026 with final shipments ending by the close of 2028, giving customers...
Tech Giants Wield Supply Chains to Crush Competition
Not just Apple making it hard on competition via supply chain power, but NVIDIA and Broadcom also. I outlined all this in my Master of the Supply Chain report. https://t.co/9ZCdGgQZVs
(PR) Solidigm Expands Sacramento Development, Fueling Global AI Leadership
Solidigm announced it has exceeded its original $100 million investment target in Greater Sacramento, now committing roughly $175 million to its Rancho Cordova headquarters and R&D campus. The company poured $75 million into a new NAND lab, added close to 100 NAND development tools,...
Memory Market Focus Shifts to Duration Over Demand
New report in the @DiligenceStack. Memory market update and sharing the framework for our tracking model of the cycle's durability. The Next Debate in Memory Is Duration, Not Demand https://t.co/gB1OsgGpBW
China's Homegrown Silicon Suppliers Gain Traction as Nvidia Struggles to Get Its Chips Into the Market — Huawei, Cambricon and...
Chinese AI and graphics chip makers have surged in 2025, capturing 41% of the domestic AI server market and cutting Nvidia's share to 55% from a claimed 95% peak. Huawei alone shipped over 812,000 AI chips, accounting for roughly half...
Nvidia Software Pushes MLPerf Inference Benchmarks To New Highs
At GTC 2026, Nvidia announced record AI inference performance on the MLPerf v6.0 benchmark, driven by its Blackwell Ultra GPUs and a refreshed software stack. The company’s Dynamo inference framework and TensorRT‑LLM optimizations delivered up to 2.77× speed gains and cut...
Bourns Adds AEC-Q Option to SSA-2 Current Sensors
Bourns announced an AEC‑Q‑compliant assembly option for its SSA‑2 analog current sensors, giving automotive, industrial and energy designers a qualified part without restarting the qualification process. The SSA‑2 series delivers precise current measurement, ultra‑low insertion loss and electrically isolated outputs...
650-V Bidirectional GaN Could Rewrite the Power-Conversion Playbook
Renesas introduced the TP65B110HRU, a 650‑V bidirectional GaN switch, at APEC 2026. The device merges a d‑mode GaN chip with silicon MOSFETs, offering 110 mΩ on‑resistance, a 3‑V gate threshold, and ±20‑V gate margin. Its true bidirectional operation enables single‑stage 500‑W solar...
Nexperia's China Unit Nears Fully Local Production of Chips: Company Sources
Nexperia’s China unit is on the cusp of achieving fully localized semiconductor manufacturing, according to internal sources. The move will enable the Dutch‑headquartered, Chinese‑owned chipmaker to produce a broader portfolio of chips within mainland China. Local production is expected to...
(PR) IBM Announces Strategic Collaboration with Arm
IBM announced a strategic collaboration with Arm to create dual‑architecture hardware aimed at enterprise AI and data‑intensive workloads. The partnership leverages IBM’s end‑to‑end system design expertise and Arm’s low‑power, scalable IP. IBM will integrate its Telum II processor and Spyre accelerator...
Nuvoton NuMicro M3331 Cortex-M33 MCU Features Built-In ARGB LED Controller, Optional USB 2.0 OTG Interface
Nuvoton has launched the NuMicro M3331 series of 32‑bit Cortex‑M33 microcontrollers running at 180 MHz, featuring a built‑in ARGB LED controller and an optional high‑speed USB 2.0 OTG interface on the M3334 variant. The devices offer up to 512 KB flash, 320 KB SRAM,...
Webinar – How to Reclaim Margin in Advanced Nodes
The ClockEdge webinar highlighted a hidden crisis in sub‑5 nm chip design: excessive guard‑banding caused by modeling uncertainty, which can strip 25‑35% of the clock period and cut performance‑per‑area (PPA) by up to 35%. Dave Johnson explained the “abstraction tax” and...
AMD Radeon RX 9070 and RX 9070 XT Fall Below MSRP in Germany
AMD’s RDNA 4‑based Radeon RX 9070 and RX 9070 XT have finally slipped below their European MSRP in Germany, with the ASUS Prime RX 9070 OC selling for €539 against a €629 list price and the ASRock RX 9070 XT Challenger at €640 versus €689. The price...
DRAM Pauses After Rally as NAND Extends Gains, Q2 Rebound Expected
DRAM prices stalled in March at $13 per 8 Gb DDR4 after an 11‑month rally, as supply agreements capped movement. Q1 pricing was largely set earlier, but quarterly contract prices still rose sharply from Q4. Samsung signaled a 40‑45% price hike...
Memory Market Rebounds, but AI Demand Reshapes Recovery
The 2026 memory market is rebounding, but the recovery is being reshaped by AI‑driven demand. TrendForce projects DRAM contract prices to climb 58‑63% QoQ and NAND Flash to surge 70‑75% QoQ in Q2. AI and data‑center expansion are pushing suppliers...
Infineon Launches TLVR Power Module for AI Server Power Demand
Infineon Technologies has launched the TDM24745T, a quad‑phase power module designed for AI server platforms. The compact 9 × 10 × 5 mm device integrates four power stages, a TLVR inductor and decoupling capacitors, achieving current densities above 2 A/mm² and peak currents up to 320 A....
Fujitsu, Rapidus Team up on 1.4nm AI Chip for Servers
Fujitsu and Rapidus announced a joint development of a 1.4 nm neural processing unit (NPU) aimed at AI inference in servers. The project, costing roughly 58 billion yen, will leverage Rapidus’s advanced node roadmap and Fujitsu’s Arm‑based Monaka CPU built on a...
Mekotronics R57-5S – Rockchip RK3576 Mini PC and Digital Player Integrates Inclined 5-Inch Touchscreen Display
Mekotronics introduced the R57-5S, a Rockchip RK3576‑based mini PC with an integrated 5‑inch inclined touchscreen aimed at kiosk and digital‑signage markets. The device offers up to 16 GB LPDDR5 RAM, up to 1 TB UFS storage, dual GbE, Wi‑Fi 5, Bluetooth 5.1, optional 4G...
Onsemi’s Hybrid Power Integrated Modules Used in Sineng Electric’s Solar and Energy Storage Solutions
onsemi announced that its next‑generation hybrid power integrated modules (PIMs) will be used in Sineng Electric’s 430 kW liquid‑cooled string energy storage system and 320 kW utility‑scale solar inverter. The FS7 IGBT and EliteSiC‑based F5BP modules deliver 32% higher power density and...
Developing A Security Framework For Chiplet-Based Systems
The article outlines a security framework for chiplet‑based systems, emphasizing that each chiplet must possess a verifiable identity tied to a platform‑wide trust chain. It describes two provisioning patterns—certificate‑based external provisioning and silicon‑derived (PUF) self‑generated keys—and explains how both feed...
Automated Multiphysics For Successful 3D-IC Design
Design teams moving to 3D‑IC architectures face intertwined power, thermal and mechanical challenges that can jeopardize yield and reliability. Traditional 2D verification tools fall short because stacked dies introduce new materials and complex inter‑dependencies. Siemens EDA’s Calibre 3DStress combined with...
AI Demand Resets Memory Market Priorities, Tightening NOR Flash Availability
The surge in AI infrastructure is creating a memory supercycle that pushes leading chipmakers to prioritize high‑margin products such as HBM, DDR5 and advanced NAND. DRAM prices are projected to jump roughly 90% quarter‑over‑quarter, while NAND could rise about 60%,...
World First: MACsec IP Receives ISO/PAS 8800 Certification For Automotive And Physical AI Security
Synopsys became the first company to earn ISO/PAS 8800 certification for its MACsec IP, a standard that secures Ethernet communication inside vehicles. The certification, validated by SGS TÜV Saar, confirms that the IP not only protects data integrity but also meets the...
Moving Electrons, Not Just Vehicles
The article examines how modern power electronics—especially multi‑level converters, silicon‑carbide (SiC) devices, and advanced power‑management ICs—are improving efficiency in electric vehicle (EV) and robot battery systems. It highlights fast‑charging challenges, noting that 15‑minute 0‑80% charges and 750 kW superchargers generate heat...
The One Bit Problem That Can Break a System
Bit flipping, once a rare reliability glitch, has become a systemic risk as semiconductor nodes shrink, clock speeds rise, and operating voltages drop, exposing aerospace, automotive and data‑center chips to silent data corruption. The phenomenon is driven by cosmic radiation,...
Embedded World 2026: Bringing Edge AI Into The Real World
At Embedded World 2026, Synaptics demonstrated that artificial intelligence is moving off the cloud and onto the device, delivering real‑time, context‑aware capabilities at the edge. The company showcased the SYN765x platform, which bundles Wi‑Fi 7, Bluetooth 6.0 and on‑chip AI compute for...
Intel Core Ultra 400HX "Nova Lake" Mobile Processor Core Configurations Surface
Intel unveiled two core configurations for its upcoming Core Ultra 400HX “Nova Lake‑HX” mobile processor, targeting high‑end gaming laptops and portable workstations. The flagship SKU packs 8 performance‑core Coyote Cove P‑cores, 16 Arctic Wolf efficiency cores and 4 low‑power island...
Altera and Arm Partner on Programmable Solutions for AI Data Centers
Altera Corp. is deepening its two‑decade partnership with Arm by integrating its data‑center‑grade FPGAs with Arm’s new AGI CPU built on the Neoverse CSS V3 architecture. The combined solution targets AI‑focused data centers, promising low‑latency, highly flexible and scalable compute...
Molex Completes Acquisition of Smiths Interconnect
Molex announced the completion of its acquisition of Smiths Interconnect, the UK‑based subsidiary of Smiths Group, marking the largest deal in Molex’s history. The purchase adds a portfolio of ruggedized connectors, RF components, optical transceivers and semiconductor‑test expertise, extending Molex’s...
SEMI Forecasts Chip Equipment Investments to Reach Beyond $150B in 2027
SEMI projects worldwide 300mm fab equipment spending to rise 18% to $133 billion in 2026 and surpass $150 billion in 2027, reaching $172 billion by 2029. The surge is driven by exploding AI chip demand for data‑center and edge workloads and by regional...
Taiwan Advances WBG Strategy Through Integrated Power Device and System-Level Innovation
Taiwan is shifting its wide‑bandgap (WBG) strategy from pure device innovation to integrated packaging, testing, and system deployment. NIKO‑SEM, a fabless power‑semiconductor designer, has broadened its portfolio from silicon MOSFETs to silicon‑carbide (SiC) and gallium‑nitride (GaN) power modules. The island...
Musk’s Terafab: Is It Real Or Not?
Elon Musk unveiled a joint Tesla‑SpaceX semiconductor fab, dubbed Terafab, slated for Austin, Texas with an estimated $20‑$25 billion investment. The plant would integrate logic, memory, and radiation‑hard chip production, targeting a 2nm process and on‑site packaging. No concrete timeline was...