AI Infrastructure

LG Energy Solution BESS Wins NVIDIA DSX Ready Qualification

Image
LG Energy Solution BESS Wins NVIDIA DSX Ready Qualification

Seoul – September 21, 2026 -- LG Energy Solution said its Battery Energy Storage System (BESS) has qualified as NVIDIA DSX Ready BESS, a designation reserved for partner products meeting NVIDIA's DSX reference design requirements for AI factory power infrastructure.

Qualification targets AI data center power bottlenecks

The DSX Ready program certifies building blocks for NVIDIA's DSX AI Factory Platform, which unifies compute, networking, power, cooling, facilities and software for AI factories. LG Energy Solution's AC-coupled BESS meets NVIDIA's applicable AC energy storage requirements, addressing AI load smoothing, voltage ride-through and power reliability challenges that data center operators face when scaling capacity.

ITRI Launches 3.2T Silicon Photonics Engine, Rallies 20+ Partners at TIE 2026

Image
ITRI Launches 3.2T Silicon Photonics Engine, Rallies 20+ Partners at TIE 2026

Taipei – September 21, 2026 -- Taiwan's Industrial Technology Research Institute (ITRI) has unveiled a 3.2 Tbps silicon photonics optical engine capable of moving roughly 400 GB of data per second, doubling the throughput of its previous 1.6T generation.

ITRI mobilizes 20+ partners to commercialize 3.2T optical technology

The technology was showcased at the Innovation Economy Pavilion of the Taiwan Innotech Expo (TIE) 2026, held at the Taipei World Trade Center Exhibition Hall 1, where the Department of Industrial Technology (DoIT) under the Ministry of Economic Affairs presented 64 technologies from more than 13 research institutes and industry partners. To push the 3.2T engine toward commercial deployment, ITRI has built a supply chain network of more than 20 partners covering chip design, fabrication, testing and packaging, and has worked with international firms to align technical specifications for the standard.

Huawei Launches SCALE System to Push Partners Toward AI at Scale

Image
Huawei Launches SCALE System to Push Partners Toward AI at Scale

Shanghai – September 21, 2026 -- Huawei unveiled its SCALE partner support system at HUAWEI CONNECT 2026, aiming to move enterprise AI deployments beyond pilot projects into full-scale industry adoption.

Huawei's Leo Chen outlines five partner challenges blocking large-scale AI rollout

Leo Chen, Huawei's Senior Vice President and President of Enterprise Sales, said during his keynote that partners face five recurring obstacles when scaling intelligent industry projects: solution development, integration verification, solution delivery, operations and maintenance, and system collaboration. The new SCALE system is structured around five corresponding support pillars: Scenario-based Solutions, Co-innovation, Aligned Marketing, Local Services with Consistent Quality, and Efficient Collaboration.

askelie Opens Operational AI Platform to MSPs and Software Vendors

Image
askelie Opens Operational AI Platform to MSPs and Software Vendors

London – September 21, 2026 -- askelie has launched its Operational AI Platform, giving managed service providers (MSPs), software vendors and business users direct access to the AI capabilities previously reserved for its enterprise deployments.

Platform lets customers configure AI tools instead of building them from scratch

The Operational AI Platform allows organisations to combine existing, reusable askelie capabilities to create AI-powered Assistants and Automations tailored to specific business needs. Askelie says the approach cuts the time, cost and risk typically associated with building an AI foundation internally.

Huawei Unveils Grid-Interactive AIDC Solution to Boost Tokens Per Watt

Image
Huawei Unveils Grid-Interactive AIDC Solution to Boost Tokens Per Watt

Shanghai – – September 20, 2026 -- Huawei unveiled a grid-interactive AI data center (AIDC) solution at HUAWEI CONNECT 2026, targeting maximum tokens-per-watt output as AI workloads strain power grids worldwide.

Huawei ties compute and energy infrastructure directly to grid stability

Hou Jinlong, Director of the Board of Huawei and President of Huawei Digital Power, opened the AIDC Facility Summit by outlining four design priorities: power acquisition, power quality, rack-level power and cooling, and faster engineering delivery. He stated that maximizing tokens per watt and minimizing cost per token will define the competitiveness of future compute factories.

Clivet Unveils Magnetic Levitation Cooling Tech for AI Data Centers

Image
Clivet Unveils Magnetic Levitation Cooling Tech for AI Data Centers

Santa Clara, Calif. – September 20, 2026 -- Clivet debuted its next-generation magnetic levitation compressor and a full data center cooling portfolio at AI Infra Summit 2026, targeting thermal management gaps created by rising AI computing densities.

Dual-rotor compressor design targets variable AI workload conditions

The new compressor uses a DUO POWER BOOST dual-rotor variable-speed architecture with two-stage compression. Two rotors and their impellers adjust speed independently to adapt to different flow rates and pressure ratios, factoring in data center load variations, water temperature conditions, and system operating strategies into cooling source design.

CleanSpark Unit Prices $2.276B Notes for Sandersville Data Center

Image
CleanSpark Unit Prices $2.276B Notes for Sandersville Data Center

Las Vegas – September 20, 2026 -- CleanSpark's wholly owned subsidiary CSDC Finance I, LLC has priced a $2.276 billion offering of 7.875% senior secured notes due 2031, sold at 98.500% of principal amount, to fund construction of the company's Sandersville Facility data center.

Notes target build-out costs, equity reimbursement and debt reserves

Net proceeds from the private offering will finance the remaining construction cost of the Sandersville Facility, reimburse CleanSpark for prior equity contributions made toward the project, and fund debt service reserves. The offering, conducted under Rule 144A and Regulation S to qualified institutional buyers and non-U.S. persons, is expected to close on September 25, 2026, subject to customary conditions.

Newsom Fast-Tracks California AI Verification Law by 8 Months

Image
Newsom Fast-Tracks California AI Verification Law by 8 Months

Sacramento – September 20, 2026 -- California Governor Gavin Newsom signed an executive order accelerating implementation of SB 813, the Fathom-sponsored law that established the nation's first comprehensive framework for independent verification organizations (IVOs) overseeing frontier AI systems.

California moves IVO designation deadline up by eight months

The executive order shifts the state's deadline to finalize IVO designation criteria from January 2028 to May 2027. SB 813 already created the legal structure for IVOs, third-party bodies tasked with independently assessing AI labs' safety claims.

Five AI Models Read Same Book, Then Critique Each Other's Analysis

Image
Five AI Models Read Same Book, Then Critique Each Other's Analysis

Paris – September 20, 2026 -- A newly published book, "Dialogue Between a Thinker and AI," has become the test object for a comparative experiment feeding identical text to five major artificial intelligence systems: OpenAI/Astra, Perplexity, DeepSeek, Google Gemini and xAI/Grok. The French and English print editions are now available, but the project's core output is a public meta-reading exercise designed to expose how differently large language models interpret the same source material.

Five AI systems receive an identical corpus for independent analysis
Each model is given the full text of the book, generated originally through hundreds of hours of dialogue between a human thinker and AI. Researchers behind the project state the goal is not to rank the models but to document where their readings converge, diverge, or miss entirely.

Dnotitia's Vector Chip Cuts CPU Load 92%, Boosts Search 5.77x

Image
Dnotitia's Vector Chip Cuts CPU Load 92%, Boosts Search 5.77x

Santa Clara – September 20, 2026 -- Dnotitia Inc. has received back the first ASIC samples of its Vector Data Processing Unit (VDPU) from fabrication, with chip-level characterization now underway and silicon-based server evaluations scheduled for the fourth quarter of 2026. The company disclosed the milestone at AI Infra Summit 2026, held Sept. 15-17 at the Santa Clara Convention Center, where it showcased a server-scale VDPU architecture built for vector retrieval workloads in retrieval-augmented generation and agentic AI systems.