
The surge in AI workloads is prompting data‑center operators to consider retrofitting legacy facilities rather than constructing purpose‑built AI campuses. Retrofitting can accelerate deployment, lower capital outlay, and improve sustainability, but it hinges on upgrading power delivery, cooling capacity, rack density, and low‑latency networking. In contrast, new AI‑centric builds guarantee optimal performance but require years of construction and significant expense. Operators must weigh the performance gap against cost, timeline, and environmental impact to choose the right path.

ElastixAI, backed by $18 million in seed funding, has emerged from stealth to launch a software platform that transforms off‑the‑shelf FPGA servers into high‑efficiency GenAI inference engines. The company argues that GPUs, designed for compute‑heavy training, are poorly suited to the...

ADLINK Technology announced two new PXI Express offerings: the PXIe-9908 eight‑channel source‑measurement unit and the PXES-2596 compact chassis. The SMU delivers four‑quadrant operation with up to 100 kS/s update and 1 MS/s sampling, aimed at high‑precision semiconductor and low‑power device testing. The...

Semtech has signed a definitive agreement with Trident IoT to embed the company’s LoRa Plus platform with Trident’s ELCap SDK, creating a royalty‑free, multi‑protocol solution for smart‑home and security devices. The partnership initially offers Z‑Wave support via the LR2021 transceiver, with...

Nvidia has not generated any revenue from the U.S.-approved H200 AI chips destined for China, despite limited approvals. The company’s Chinese data‑center sales once represented roughly 20% of its revenue, but export controls forced it to create a lower‑performance H20...

Vertiq has launched servo firmware version 0.1.1, extending DroneCAN and IQUART functionality across its G2 and G1 servo modules. The update adds ArrayCommand support, ActuatorStatus telemetry, and dynamic node‑ID features while introducing unitless control types and linear displacement mapping for...

Nvidia CFO Colette Kress told analysts that the current shortage of graphics chips and cards for gaming rigs will persist for the next few quarters. The tight supply stems from Nvidia’s aggressive allocation of fab capacity to higher‑margin AI accelerators....
SK Hynix announced an additional 21.6 trillion won ($15 bn) investment to expand its first semiconductor fab in Yongin, raising the total spend to 31 trillion won. The funding, slated for March 2026‑December 2030, will add five clean rooms, bringing the site to six and...

Nio’s intelligent‑driving chip subsidiary, GeniTech (Shenji), has signed definitive agreements to secure a CN¥2.26 billion investment from a consortium of Chinese investors. The investors will purchase newly issued shares, giving them a collective 27.3% stake while Nio retains a 62.7% controlling...

Nvidia announced that its Shield TV line will continue receiving software support, rolling out the Shield Experience 9.2.4 update for all devices. The patch brings January 2026 security updates and resolves several high‑profile bugs, including Disney+ playback and Bluetooth remote disconnects. This...

LM Studio and Tailscale have launched LM Link, a feature that lets developers access remote GPU rigs as if they were locally attached. The solution replaces public APIs and SSH tunnels with a private, WireGuard‑encrypted tunnel built on Tailscale’s userspace tsnet...
A user on an enthusiast forum is attempting chip‑level repair of a Biostar B460GTA Rev 5.0 motherboard but cannot locate the boardview or circuit diagram. The poster, based in China, sought help downloading a purported schematic from a dubious website and...

Mitsubishi Heavy Industries has launched DIAVAULT, an industrial‑grade edge data centre platform that integrates power, cooling and AI compute for on‑premise processing. The solution targets low‑latency, secure workloads in manufacturing, defense and 5G environments, addressing data‑privacy and latency concerns of...

Nvidia projected fiscal Q1 revenue of about $78 billion, topping the average Wall Street forecast of $72.8 billion and approaching the high‑end $80 billion outlook. CEO Jensen Huang highlighted a surge in enterprise adoption of AI agents, indicating that corporate customers are joining...
Quobly, a French silicon‑based quantum computing pioneer, has signed a Memorandum of Understanding with Singapore’s Entropica Labs to co‑design fault‑tolerant quantum systems. The alliance integrates Entropica’s compiler and optimization stack directly with Quobly’s silicon spin‑qubit architecture, targeting early‑stage hardware‑software co‑design....

AMD announced a $250 million strategic partnership with Nutanix, buying $150 million of Nutanix common stock and committing up to $100 million for joint R&D and go‑to‑market efforts. The collaboration will integrate Nutanix’s Cloud and Kubernetes platforms with AMD’s EPYC CPUs, Instinct GPUs,...
Terawulf Inc. reported a transformative 2025, shifting from Bitcoin mining to a power‑backed high‑performance computing (HPC) platform. Revenue rose 20% to $168.5 million, driven by $12.8 billion of credit‑enhanced HPC lease contracts and $6.5 billion of new financing. Adjusted HPC margin reached 77%...
D‑Wave Quantum reported fiscal 2025 revenue of $24.6 million, up 179% year‑over‑year, driven by $16.2 million in system sales and $5.5 million in QCaaS subscriptions. Bookings jumped 471% in Q4 to $13.4 million after a €10 million multiyear contract, while the sales pipeline expanded nearly...

Dell introduced the PowerEdge XR9700, a ruggedized, liquid‑cooled server built for edge AI and cloud‑RAN deployments in harsh outdoor settings. The chassis is IP66 sealed and operates from –40°F to 115°F, eliminating the need for climate‑controlled shelters. Powered by Intel’s...
Renishaw has integrated time‑resolved Raman spectroscopy (TRRS) into its inVia confocal Raman microscope, pairing a newly developed single‑photon avalanche diode (SPAD) array detector with 50‑picosecond timing precision. The technique isolates Raman‑scattered photons from slower fluorescence, delivering high‑quality spectra for samples...

Nvidia posted a record $68 billion quarter, a 73% year‑over‑year increase, driven largely by its data‑center business. Data‑center revenue reached $62 billion, split between $51 billion in compute and $11 billion in networking products, pushing full‑year revenue to $215 billion. The company still reports no...

Motorola unveiled a preview of its upcoming Razr Fold at CES 2026, showcasing a surprisingly thin, book‑style foldable that sits between the Galaxy Z Fold 7 and Pixel 10 Pro Fold in design. The company hinted that full specifications and a potential launch timeline will...
Ascent Solar Technologies has supplied its lightweight, flexible CIGS photovoltaic blankets to NOVI Space’s N‑1 ATLAS spacecraft. The N‑1 ATLAS, a toaster‑sized hyperspectral imaging satellite, will launch aboard SpaceX’s Falcon 9 Transporter‑16 in March and generate 150 W of power. The integration...
![Samsung’s Audio Eraser Now Works with YouTube and Other Apps, and It’s Impressive [Video]](/cdn-cgi/image/width=1200,quality=75,format=auto,fit=cover/https://i0.wp.com/9to5google.com/wp-content/uploads/sites/4/2026/02/galaxy-s26-audio-eraser-2.jpg?resize=1200%2C628&quality=82&strip=all&ssl=1)
Samsung has upgraded its Audio Eraser tool on the Galaxy S26, enabling real‑time voice isolation across popular streaming and social apps. Integrated into One UI 8.5, users can toggle the feature, adjust its strength, and enable a Voice Focus mode. Demonstrations show...

The FBI’s Criminal Justice Information Services division has issued a Request for Information seeking near‑infrared iris cameras that can capture both eyes in a single enrollment. The RFI covers fixed devices for booking stations and mobile units for field use,...

The article shows how to transform an outdated Android phone into a functional e‑reader with a few simple tweaks. By swapping the stock launcher for a minimalist alternative, enabling a blue‑light filter, and loading reading‑focused apps, the device mimics the...

Vast Data and Nvidia have launched the CNode‑X, a GPU‑powered server that embeds the Vast Data AI Operating System directly onto Nvidia hardware. The integrated solution is optimized for AI pipelines, high‑performance analytics, vector search, retrieval‑augmented generation and agentic workloads....

Mouser Electronics, in partnership with Microchip Technology and Samtec, has released an eBook titled “8 Experts on PCIe for Emerging Embedded Systems.” The guide gathers insights from eight industry leaders—including Bosch, DANA, Synopsys, Tektronix and Visteon—to help engineers tackle PCIe integration...

The article explains why 15,000 RPM hard‑disk drives never achieved mainstream adoption. Extreme rotational speeds caused heat, power draw, vibration‑induced head crashes, and intolerable acoustic noise. These mechanical barriers made the drives costly and unreliable, especially compared with emerging solid‑state technology....

Internet Service Providers shape every click by routing traffic through dozens of networks, adding latency that spikes during peak hours. Congestion can degrade residential speeds by 25% or more, especially where legacy copper lines remain. ISPs also employ traffic‑shaping tools...

HPE’s AI and supercomputing leader Nithin Mohan argues that enterprise AI is now limited by infrastructure rather than algorithms. He highlights how exascale computing, high‑speed data movement, and system reliability are essential to move AI from demos to production. The conversation...
GlobalWafers has launched Phase II production at its Sherman, Texas silicon‑wafer fab, a $4 billion expansion that brings total regional investment to $7.5 billion. The 12‑inch wafer line will supply Texas Instruments, TSMC and other partners producing chips for Apple. The project receives...
A new review in Nanoenergy Advances highlights how atomically thin, graphene‑like memristors can store information directly in their molecular lattice, offering fast, dense, and energy‑efficient switching. The paper details how graphene oxide, diamane, and layered chalcogenides achieve controllable resistance changes...

Rohde & Schwarz and LITEON demonstrated a production‑optimized test setup at MWC Barcelona 2026, using the PVT360A vector performance tester to characterize four FlexFi 5G femtocells in parallel. The PVT360A’s 2 × 8‑port architecture and smart‑channel technology cut testing time by 50 percent, enabling fully automated...
ESAB, together with Linde Gas & Equipment, unveiled the Ruffian EMP 270G engine‑driven welding generator at CONEXPO‑CON/AGG 2026, offering show‑only pricing and bundled accessories. The unit delivers 270 A at a 100 % duty cycle and 11 kW of generator power from a 23.5‑HP KOHLER...

AI‑ready Aitech SP1 single‑board computer now incorporates Teledyne e2v’s QLS1046 space‑qualified SoC. The chip features four 64‑bit Arm Cortex‑A72 cores, DDR4 memory with error correction, and radiation tolerance up to 100 krad, enabling reliable operation in LEO, GEO, lunar and deep‑space...

The NDSS 2025 paper introduces Borrowed Time, a countermeasure that protects integrated circuits from emerging static side‑channel attacks such as static power analysis, laser logic state imaging, and impedance analysis. By continuously monitoring a device and securely erasing key‑dependent data...

Microsoft clarified that Windows 11 will not drop support for legacy V3 and V4 printer drivers, countering earlier reports. Existing drivers remain functional and available via Windows Update, while new driver submissions are blocked by default and require case‑by‑case approval starting...

Multimatic Motorsports completed a full‑distance endurance race using Conflux Technology’s 3D‑printed, configurable transmission oil cooler. The additive‑manufactured unit was produced in two weeks and delivered roughly 20% more heat rejection than the legacy cooler while fitting within the same compact...

DigiKey will showcase more than a dozen new embedded‑system products from leading suppliers at Embedded World 2026 in Nuremberg, Germany. The booth (4A‑633) features live demos ranging from thread‑network sensor arrays to single‑pair Ethernet video streaming and AI‑enabled microcontrollers. DigiKey...

China plans to quintuple its leading‑edge chip output within two years, raising 7 nm and 5 nm wafer starts from under 20,000 to about 100,000 per month. The long‑term goal is to reach 500,000 advanced‑node wafers monthly by 2030. SMIC, the only...
Researchers have demonstrated lead‑free all‑inorganic double perovskite Cs2AgBiBr6 memristors that combine robust analog resistive switching with biologically relevant synaptic plasticity. The devices rely on reversible Ag⁺ ion migration to form filamentary conduction paths, as confirmed by impedance spectroscopy, c‑AFM, and...

MathWorks announced its membership in the EDGE AI Foundation, a consortium dedicated to energy‑efficient artificial intelligence for edge devices. The partnership will let engineers leverage MATLAB and Simulink to train, integrate, and deploy AI models on embedded hardware while validating...

Stratasys announced a 22% year‑over‑year reliability increase for its F900 industrial 3D printer, driven by manufacturing enhancements co‑developed with its Customer Advisory Board (CAB). The CAB, composed of aerospace, defense and automotive leaders, identified production‑scale barriers and helped shape improvements...

Scality and WEKA announced that Scality RING will serve as the back‑end object store for WEKA’s NeuralMesh high‑performance AI file system. The partnership leverages NeuralMesh’s SSD‑based front‑end with RING’s cost‑efficient, disk‑based object tier, delivering up to ten times faster performance than...
STIGA S.p.A. has partnered with SiMa Technologies to embed SiMa.ai’s low‑power Physical AI ML‑SoC into its next‑generation robotic lawn mowers for both consumer and commercial use. The integration enables ultra‑low‑latency, edge‑based decision‑making while reducing power draw, resulting in lighter, more...

AutomationDirect announced the addition of Penn‑Union split bolt connectors, aluminum power bars, and bronze service post connectors to its catalog. The split bolts accommodate CU‑CU, CU‑AL, and AL‑AL conductor pairs, while the power bars support two to four conductors in...
Japanese tech giant Toshiba and MIRISE Technologies have embedded their Simulated Bifurcation Machine, a quantum‑inspired optimization engine, into an autonomous mobile robot. The on‑board FPGA runs a new multi‑object tracking algorithm at 23 frames per second, delivering up to 23%...
Mersen has introduced its Critical Power Panel (CPP) series, a fully customizable fused panelboard that combines a lightweight design with a reduced footprint. Targeted at NEC‑required and non‑required applications—including UPS systems, data centers, medium‑voltage eHouses, emergency circuits, and DC utility...

Dual Inventive has completed field trials of its next‑generation rail safety suite, including the RCS 3000 remote‑control signal, the RSS 3000 PC off‑grid points controller, and the RDI 3000 v2 remote detonator isolator. Tests in Germany, France, Belgium and the UK demonstrated clear visibility...