Today's Hardware Pulse
Nvidia driver hints at next‑gen DLSS 5 work
Nvidia’s GeForce Game Ready driver 610.47 adds three new profile options—Enable DLSS NR override, Enable DLSS NR SL override, and Override DLSS NR presets—suggesting continued development of DLSS 5 neural‑rendering technology. The earlier DLSS 5 reveal had sparked debate over AI‑driven image alteration, and the new driver hints at a quiet but ongoing effort.
Also developing:
By the numbers: Cyient Semiconductors raises $30M Series A
FreshAI at Checkout: Faster Exits. Happier Returns.
Mettler Toledo’s FreshAI at Checkout uses AI‑driven item recognition to speed grocery checkout lines, delivering 99% accuracy and scanning each item in roughly half a second. The solution eliminates manual entry errors, cuts labor costs, and reduces shrinkage, translating into higher throughput and faster transactions. Its SDK‑based integration fits any POS system without additional infrastructure, enabling rapid store‑wide rollout. Czech retailer Albert’s recent RETA 2026 AI award highlights the technology’s market traction and proven impact on shopper satisfaction and profitability.
Making Waves
Microwave Journal released its May 2026 "Making Waves" issue, a downloadable PDF that spotlights the latest breakthroughs in microwave and millimeter‑wave technology. The publication bundles expert commentary, case studies, and data on emerging 5G, satellite‑comm, and advanced material applications. By providing...
New Stair Evacuation System Quickly Safely Moves Patients
Emergency Stair Evacuation System: Safely Drag Patients Downstairs in Seconds by @IntEngineering #Innovation #EmergingTech #TechForGood #Technology #Tech https://t.co/ZuwCdyINJx
Logic Scaling Boosts Efficiency, Yet Costs Rise Sharply
Scaling logic continues to deliver better performance per watt, but it’s becoming harder, more expensive, and increasingly customized. https://t.co/MzILzSSFrN #2nm #semiconductor
JEDEC Releases New Memory Interface Logic and Expanded MRDIMM Roadmap
JEDEC’s JC-40 and JC-45 committees announced the release of the DDR5MDB02 multiplexed rank data buffer standard and signaled an imminent DDR5MRCD02 clock‑driver specification. The groups are also finalizing the MRDIMM Gen2 module standard and prototyping raw‑card designs that reach 12,800 MT/s....
Practical Failure Modes in Remote IO Systems
The article outlines the most common failure modes in remote I/O systems, including network communication loss, power‑supply instability, harsh environmental exposure, and neglect of built‑in diagnostics. It explains how each issue manifests—timeouts, voltage dips, component drift, or silent faults—and why...
Indoor Positioning Moving Toward Layered, Multi-Radio Architectures
The indoor‑positioning market is moving from a single‑technology race to a layered, multi‑radio architecture that blends Wi‑Fi 802.11az, ultra‑wideband (UWB) and Bluetooth channel sounding. Wi‑Fi is gaining ranging capability through time‑of‑flight measurements, while UWB remains the gold‑standard for 10‑30 cm precision...
AMD Leaks First 16‑Core Ryzen 9 PRO 9965X3D with 3D V‑Cache
AMD has leaked specifications for the Ryzen 9 PRO 9965X3D, its first 16‑core processor in the PRO portfolio and the inaugural PRO chip to carry 3D V‑Cache. Built on the Zen 5 architecture, the chip targets professional, content‑creation and AI workloads while promising...
Nvidia Adds Taiwan's JPC as Second Certified Supplier for Vera Rubin AI Power Cables
Nvidia confirmed that Taiwan‑based JPC Connectivity has become its second certified supplier of power cables for the Vera Rubin AI supercomputer platform. The move reflects growing demand for high‑power components as AI workloads surge.
Patna Deploys AI‑Powered “Nagar Netra” Fleet to Auto‑Fine Littering, $1.1M Smart‑City Rollout
Patna Smart City Limited has launched an AI‑driven camera system on 19 electric two‑wheelers that will patrol all 75 municipal wards, automatically detecting garbage piles and issuing fines. The Rs 8.95 crore ($1.1 million) project aims for full deployment by November 2026 and signals...
BrainChip Strikes IP Licence Deal with ASICLAND to Integrate Akida AI Technology
BrainChip Holdings (ASX:BRN) has signed a non‑exclusive, worldwide IP licence agreement with ASICLAND to embed its Akida neuromorphic AI technology into ASICLAND’s system‑on‑chip designs. The deal provides upfront evaluation and production licence fees per customer plus volume‑based royalties on net...
From AI Servers to High-Speed PCBs: Why CCL Prices Are Entering a New Upcycle
Copper‑clad laminate (CCL) prices are entering a new upcycle as high‑frequency, AI‑driven servers, 5G infrastructure, and electric vehicles demand faster, thinner, and greener boards. Environmental mandates for lead‑free and halogen‑free formulations add complexity, while tighter dimensional tolerances push material costs...
Denso Aims to Build System to Wirelessly Charge Moving EVs by FY 2029
Japanese auto parts supplier Denso, a Toyota affiliate, announced it aims to commercialize a wireless charging system that powers electric vehicles while they are in motion by fiscal 2029. In a September 2024 test at its Aichi headquarters, the prototype...
Vertiv Posts 83% Earnings Jump on $15 B AI Data‑Center Backlog
Vertiv announced an 83% rise in earnings, crediting a $15 billion AI data‑center project backlog that is spurring demand for its liquid‑cooling solutions. The surge lifts adjusted operating margins to 20.8% and underpins a bullish outlook for the next 18 months.
Jefferies: AI Capex to Consume 92% of Hyperscalers' Cash Flow, Powering DRAM Surge
Jefferies projects that the four biggest U.S. hyperscalers will allocate 92% of their operating cash flow to capital spending on AI infrastructure in 2026, up from 41% in 2023. The surge, part of a $800 billion AI‑related capex wave, is channeling...
Australian Team Demonstrates Quantum Battery That Charges Instantly and Scales with Size
Researchers from CSIRO, the University of Melbourne and RMIT have built the world’s first functional quantum battery prototype that charges in a single rapid burst and charges faster as its size increases. The breakthrough, published in Nature Light: Science &...
Tata Electronics Aims to Be $30 Billion Business with Fab Play: CEO & MD Randhir Thakur
Tata Electronics, the Tata Group’s semiconductor arm, has surged from a ₹400 crore base to a ₹1.3 lakh crore (≈$15 bn) revenue run‑rate in four years and now aims to double that to $30 bn by 2031. The company is building India’s first fab in...
Samsung Galaxy Z Flip 8: Insider Comments on Price and Upgrades
Samsung is set to launch the Galaxy Z Flip 8 at its July 2026 Unpacked event, introducing a new Exynos 2600 processor, a redesigned hinge that makes the device 0.02 inches thinner when folded, and a crease‑free display. The phone retains the same...
Adiuvo Explorer Board Aims to Bring Artix UltraScale+ FPGA to $99 Platform
Adiuvo is engineering the Explorer Board, a $99 FPGA development platform built around Xilinx’s Artix UltraScale+ AU7P. The board packs 37K LUTs, 216 DSP slices, 3.6 Mb of block RAM and integrated GTH transceivers, and adds external HyperRAM for larger memory...
Doubling Throughput with Speculative Decoding on DGX Spark
Pretty sure I can double the t/s Isn’t possible to do speculative decoding with a DGX spark on the same network?
12‑Sensor Glove Offers Ultra‑Precise VR Hand Tracking
Glove-Type #VR Controller: 12 #Sensors and Haptic Feedback for Ultra-Precise Hand Tracking via @ZappyZappy7 #AR #AugmentedReality #VirtualReality #TechForGood https://t.co/gK3mXZQaIb
PsiQuantum Appoints Intel CEO Lip‑Bu Tan to Board, Sharpening Quantum Hardware Roadmap
PsiQuantum announced that Intel CEO Lip‑Bu Tan has joined its board of directors, bringing decades of semiconductor leadership to the quantum‑computing startup. The move coincides with a reshuffle that placed former AMD president Victor Peng as interim CEO and co‑founder...
US Cloud Giants Commit $700 Billion to AI Infrastructure, Sparking ASIC Demand Uncertainty
Amazon Web Services, Microsoft Azure, Alphabet’s Google Cloud and Meta’s data centers together pledged roughly $700 billion in AI‑focused capital spending for 2026. While Alphabet disclosed a $190 billion AI budget, analysts warn that the surge could outpace ASIC supply, leaving timing...
Your iPhone Runs on Oil
The iPhone’s supply chain is deeply intertwined with petroleum at every stage, from diesel‑powered mining of copper, lithium and rare earths to the high‑heat smelting processes that turn ore into usable metal. Petrochemical feedstocks create the plastics, adhesives, solvents and...
OnePlus Launches Strix G15 Gamepad to Turn Ace 6 Ultra Into a Handheld Console
OnePlus introduced the Strix G15 gamepad in China, a metal‑framed accessory that docks the Ace 6 Ultra tablet and delivers sub‑50 ms Bluetooth latency. Priced at 599 yuan (about $84), the controller aims to capture a fast‑growing mobile‑gaming accessories market.
Semis Surge, Stocks Pricey, yet Indices Stay Calm
Semis are up 30%+ in a month, single stock IVs are quite rich, with the Iran conflict lingering in the background…yet major indices are seeing the lowest vol in months. Read our weekly analysis at: https://t.co/jhbxoVyE7x
DJI Drone Sales Banned in Beijing and US Over Security Fears
DJI's flagship drones are now prohibited from sale in Beijing and blocked from new US authorisation after the Chinese capital and the FCC cited security risks. The bans threaten $1.5 billion of projected US revenue and could curtail the company's global...
Boardcon PICOT536 SoM and EMT536 SBC Feature Allwinner T536 Edge AI Processor
Boardcon has introduced the PICOT536 system‑on‑module and the EMT536 development board, both built around Allwinner’s T536 edge‑AI processor. The T536 combines a quad‑core Cortex‑A55 CPU, RISC‑V co‑processors and a 2 TOPS NPU, supporting up to 8 GB LPDDR4/LPDDR4X and 64 GB eMMC storage....
Azure Updates: Integrated HSM; GPT-5.5; API Management; Discovery
Microsoft unveiled Azure Integrated HSM, a Microsoft‑built hardware security module that meets FIPS 140‑3 Level 3 standards and is now available as open‑source firmware, software, and drivers. At the same event, OpenAI’s GPT‑5.5 entered general availability through Microsoft Foundry, offering multi‑step reasoning,...
FranklinWH Introduces Enhanced 15 kWh aPower in Australia and New Zealand
FranklinWH launched an upgraded 15 kWh aPower battery in Australia and New Zealand, expanding usable capacity from 13.6 kWh to 15 kWh and boosting warrantied throughput by 40% to 60 MWh. The enhancement keeps the same footprint and price while promising up to $3,900 in...
The US Has Banned the World’s Best Drones. It Has Not Figured Out How to Make Them.
The FCC’s December 2025 decision automatically placed DJI on the Covered List, halting imports of new DJI drones that command roughly 80% of the U.S. consumer and commercial market. In response, Skydio unveiled a $3.5 billion, five‑year SkyForge program to build a...
Best Record Players I've Tested for 2026
CNET evaluated a range of turntables, from budget-friendly models like the Audio‑Technica AT‑LP60X to premium options such as the U‑Turn Orbit Theory and Rega Planar 3. Scores ranged from 8.3 to 8.6, highlighting strengths in sound quality, automatic operation, and modern...
Disabling EXPO and PBO Adds Thousands of Benchmark Points
EXPO: off EXPO High Bandwidth: Disabled PBO: Disabled I was running my PC like it was an i5 Ran some benchmarks after setting everything and it improved a few thousand points. Now I'll experiment with 90 Level 2
Apple’s First Wearable: A Plain Watch in 1995
#ThisDayInTechHistory. May 2, 1995. @Apple released it's first wearable. A watch. Just a watch. No web connectivity, no notifications, no fitness tracking just a watch. #Wearables https://t.co/pVthKdKTe1
Self-Healing Synaptic Transistor Recovers Memory After Damage
Researchers have created a fully self‑healing, stretchable synaptic transistor that regains most of its function after being cut in half. The device restores about 80% of its operating current and over 90% of its memory within 24 hours without external triggers,...
Nvidia's China AI Accelerator Share Hits 0% Amid Export Controls
JUST IN: Jensen Huang says Nvidia's market share of AI accelerators in China has dropped to 0%, calling U.S. export controls strategically counterproductive.
Low GPU Use Beats Hyperscalers, Accelerating Multi‑Tenant Optimizations
GPU utilization is embarrassingly low. This is actually good versus the hyperscalers. We’re about to speedrun the multi-tenant CPU optimizations of the last 25 years, including all the security headaches, but with GPUs.
The Steam Deck OLED Beats Every Rival because of One Thing No Windows Handheld Can Match
Valve’s Steam Deck OLED, released late 2023, stands out with a 7.4‑inch 720p OLED panel that reaches 1,000 nits and runs at 90 Hz, delivering vivid HDR performance. Its Linux‑based SteamOS, bolstered by the Proton compatibility layer, runs thousands of Windows...
Apple’s “Pro” Label: MacBook vs Mac Pro Complexities
Apple calls a lot of its products "Pro." How do the MacBook Pro and Mac Pro compare? It's complicated. https://t.co/xEyw97cqUr
3D‑Printed Robotic Hand Replicates Human Gestures Instantly
#3DPrinted #Robotic Hand Mirrors Human Gestures in Real Time by @YKwolfpec #EmergingTech #Innovation #TechForGood #Tech https://t.co/BYV0ks5vzO
South Korea’s April Exports Surge 48% to Record $85.9 B, Driven by AI‑Fueled Chip Boom
South Korea’s exports rose 48% year‑over‑year in April, reaching $85.9 billion, the second‑highest monthly total on record. The surge was powered by a 173.5% jump in semiconductor shipments tied to artificial‑intelligence demand, while auto exports slipped amid Middle‑East logistics disruptions. The...
India Forecasts Six‑Fold Data‑Center Expansion to 10.5 GW by FY31, Triggering $60 Billion Capex
Morgan Stanley projects India's data‑center capacity will rise nearly six‑fold to 10.5 GW by FY31, driven by AI workloads and data‑localisation rules. The surge will demand about $60 bn in capital spending, including $20 bn for power and renewable upgrades, reshaping the country's...
Argonne Lab Shows Electron‑on‑Neon Qubits Cut Noise Up to 10,000‑Fold
Researchers at Argonne National Laboratory and the University of Notre Dame have demonstrated that electron‑on‑neon qubits exhibit noise levels 10 to 10,000 times lower than traditional semiconductor qubits. The findings, published in Nature Electronics, could reshape the race for scalable...
Firestarter Malware Persists on Cisco ASA/Firepower Devices After Patches
U.S. CISA and U.K. NCSC have disclosed that the custom Firestarter malware can survive Cisco ASA and Firepower firmware updates, giving threat actors continued access to compromised firewalls. The backdoor exploits CVE‑2025‑20333 and CVE‑2025‑20362 and hooks into the LINA process,...
5 Free Fixes that Make Every Smart TV Perform Like a Premium Model
The How‑To‑Geek guide outlines five free tweaks that turn any smart TV into a near‑premium display. It advises disabling motion‑smoothing features such as Samsung’s Auto Motion Plus, switching to a wired 10/100 Mbps Ethernet connection, and selecting Filmmaker or Movie picture...
1983 Debate: Home Computers Predicted to Last
Predicting the future is hard work. In 1983, most people were betting that the home computer was a passing fad. Two @MIT professors debated the future of PCs in 1983. Prof. Michael Dertouzos tried to convince us that the home computer...
This $19.99 Roku Camera Turns Any Room Into a Budget-Friendly Smart Security Setup
Roku’s Indoor Camera is now priced at $19.99 on Amazon, a 33% discount from its $29.99 list price. The device delivers 1080p HD video, full‑color night vision, motion and sound alerts, two‑way audio, and Wi‑Fi 6 connectivity. Integrated with Roku TVs...
New Wafer Inspection and Metrology Platform From TRI
TRI unveiled the AI‑powered TR7950Q SII wafer metrology and inspection platform, targeting back‑end process and advanced‑packaging applications. The modular system handles 6‑inch to 12‑inch wafers on a high‑stability granite base and offers automated visual inspection for particles, scratches, and contamination....
Selling Older GPUs to China Boosts US Security
The argument for selling deprecated GPUs to China: Selling them less advanced GPUs than we have in America decreases the odds that they develop more advanced GPUs than America that go down a different, more power intensive tech tree. Pro national-security.
Isolated 15 W and 30 W Converters Cut Complexity
TDK‑Lambda has introduced the CCGS series, a line of isolated DC‑DC converters delivering 15 W or 30 W in compact chassis‑mount (52 × 93 × 23.5 mm) and DIN‑rail (52 × 93 × 31.9 mm) formats. The converters accept 9‑36 V or 18‑76 V inputs and provide selectable single or dual outputs from 3.3 V...