AI Infrastructure & Data Centers

SK hynix Launches Venture Arm to Fund Global AI Ecosystem

Seoul – September 20, 2026 -- SK hynix Inc. has launched a new corporate venture capital brand, SK hynix Ventures, to expand strategic investments across the global AI ecosystem, the company announced at the inaugural 'SK hynix Ventures Day' held in Silicon Valley.

CEO Kwak Noh-Jung leads executive delegation at Silicon Valley launch event

CEO Kwak Noh-Jung and senior SK hynix executives met with representatives from global venture capital firms and leading startups to outline investment strategies targeting AI data centers and optical-based systems. The event brought together startups from the U.S. AI computing and data center sectors to discuss memory, computing and system software innovations tied to advanced AI models.

Arcjet Launches Runtime Security Product for Production AI Agents

San Francisco – – September 20, 2026 -- Arcjet has launched agent runtime security, a new product designed to give engineering and security teams visibility and control over AI agents operating inside production systems. The platform targets agents that read and write to databases, process refunds, call APIs, and take autonomous actions across multiple systems on behalf of users.

Arcjet builds its product around three functions: observe, enforce, and audit

The observe layer lets teams ingest agent activity without changing application code, using existing OpenTelemetry tooling or, for Claude users, the Claude Compliance API. Arcjet connects actions across sessions into a single workflow view, capturing prompts, tool call parameters, session metadata, identity, and security decisions, and maintains an inventory of agents and applications running in a given environment.

PrismML's Bonsai 2 27B Cuts AI Memory Use 9x, Keeps 98% Accuracy

Pasadena, Calif. – September 20, 2026 -- PrismML has released Bonsai 2 27B, a ternary-compressed version of a 27-billion-parameter multimodal model that runs at just 5.9 GB while retaining 98.2% of its full-precision counterpart's benchmark performance.

Ternary compression shrinks memory footprint by more than 9x

Built on Qwen3.8 27B, Bonsai 2 27B reduces the model's memory requirement to 5.9 GB, a reduction of over nine times compared with the full-precision original. The model contains 27.8 billion parameters and was trained using Google v5 TPUs.

Delta Electronics Speeds NVIDIA DSX AI Factory Rollouts with 800 VDC Tech

Image
Delta Electronics Speeds NVIDIA DSX AI Factory Rollouts with 800 VDC Tech

Fremont – September 20, 2026 -- Delta Electronics announced a system-level infrastructure approach to accelerate deployment of AI factories built on the NVIDIA DSX AI Factory Platform, integrating power delivery from onsite energy sources down to the chip level to raise token output per megawatt.

Delta's 800 VDC architecture cuts conversion stages and hits 98% power efficiency

The company's 800 VDC power architecture reduces conversion stages between facility power and compute, delivering up to 98% power conversion efficiency while responding to GPU load transients. High-density In-Row systems can deliver up to 800 kW in a single power rack. At the board level, Delta's 800 VDC DC-DC distribution technology steps down directly to 50 VDC or 12 VDC with peak efficiency reaching 98.5%.

WiMi Unveils Reinforcement Learning Framework for Quantum Encoding Circuits

Beijing – September 20, 2026 -- WiMi Hologram Cloud Inc. (NASDAQ: WiMi) has developed a reinforcement learning-based framework to automate the design of quantum encoding circuits, aiming to cut the resource costs and inefficiencies that have limited quantum machine learning (QML) model performance.

WiMi replaces manual circuit design with automated search architecture

The new scheme uses reinforcement learning to generate task-specific encoding circuit architectures automatically, moving away from the heuristic, manually designed approaches that have dominated QML development. WiMi states the framework targets three persistent industry problems: poor task adaptability, low search efficiency, and insufficient handling of multiple design objectives simultaneously.

Advantech Adds Intel, Qualcomm, AMD to WEDA Edge AI Ecosystem

Image
Advantech Adds Intel, Qualcomm, AMD to WEDA Edge AI Ecosystem

Taipei – September 20, 2026 -- Advantech Co., Ltd. (TWSE: 2395) has expanded its WEDA-Powered Edge AI ecosystem by bringing Intel, Qualcomm Technologies, and AMD into a shared framework for developing and deploying Edge AI across industrial and enterprise hardware.

Advantech unifies Edge AI development under WEDA architecture

The expansion centers on WISE-Edge Developer Architecture (WEDA), WEDA-Ready Edge Computing, and the Advantech Container Catalog (ACC). Together, the components are designed to standardize Edge AI development, validation, deployment, and lifecycle management across fragmented hardware environments, replacing what Advantech describes as disconnected, project-by-project development efforts.

Huawei Unveils Grid-Interactive AI Data Center Solution to Boost Tokens Per Watt

Image
Huawei Unveils Grid-Interactive AI Data Center Solution to Boost Tokens Per Watt

Shenzhen – September 20, 2026 -- Huawei unveiled a grid-interactive AI data center (AIDC) solution at the Green Energy Development Conference & International Digital Energy Expo 2026, targeting power acquisition, high-density cooling, and delivery bottlenecks as AI campuses scale from megawatts to gigawatts.

Huawei's Zhou Jianjun says AIDC load fluctuations now directly impact grid operations

Zhou Jianjun, Vice President of Huawei and President of Global Marketing, Sales and Services at Huawei Digital Power, presented the solution during the conference's key industry achievement release session. He stated that startup, shutdown, and load fluctuations of large AIDC campuses can affect grid operations, and that declining local grid strength combined with rising renewable penetration is forcing data centers to become "grid-friendly" nodes rather than passive consumers.

87% of Firms Waste Hours Re-Verifying AI Agent Context, Survey Finds

Image
87% of Firms Waste Hours Re-Verifying AI Agent Context, Survey Finds

New York and Brussels – September 20, 2026 -- Nearly nine in ten enterprise decision-makers (87%) say their teams routinely burn hours re-verifying that context fed to autonomous AI agents remains accurate, according to a Harris Poll survey of 306 U.S. data, privacy, and AI decision-makers commissioned by Collibra. The report, titled The 2026 Hallucination Tax Report, found that 76% of organizations have hit critical roadblocks trying to move AI agents from pilot programs into full production over the past 12 months.

CAEVES Taps Ex-NetApp Technologist to Expand AI Data Platform Beyond Copilot

Image
CAEVES Taps Ex-NetApp Technologist to Expand AI Data Platform Beyond Copilot

New York – September 19, 2026 -- CAEVES Technology, Inc. has appointed Kirk Ryan, a two-decade infrastructure veteran, as Director of AI Innovation on its Advisory Board, tasking him with extending the company's AI Data Platform beyond Microsoft Copilot to additional AI assistant ecosystems.

Ryan brings NetApp and consultancy pedigree to CAEVES' AI buildout

Ryan previously served as Global Principal Technologist at NetApp and is Managing Director of KDBL, a UK-based AI consultancy. He is known for commentary on enterprise AI, retrieval-augmented generation, search, and the Model Context Protocol.

Trackonomy Acquires Identiv's IoT Label Unit, Forms Trackonomy Group

Image
Trackonomy Acquires Identiv's IoT Label Unit, Forms Trackonomy Group

San Jose, Calif. – September 19, 2026 -- Trackonomy Systems has completed the acquisition of Identiv's IoT label business, creating Trackonomy Group, a vertically integrated Physical AI supply chain spanning semiconductors, smart labels, and end-to-end tracking solutions.

Three independent businesses now orchestrate nearly 20 million shipments daily

The newly formed group combines InPlay, a Bluetooth Low Energy semiconductor supplier; Identiv, a manufacturer of NFC-, RFID-, and BLE-enabled IoT labels; and Sentient, the original smart label pioneer with readers, label-application robotics, camera tunnels, and a cloud platform. Together, the three units process nearly 20 million shipments daily and are targeting more than 200 million connected assets.