The Ultimate Guide to AI Infrastructure
A curated Kiwi edition of IndustrialDay news, analysis, interviews, reviews, job moves, and related resources for AI Infrastructure.
What to know about AI Infrastructure
AI Infrastructure explores the hardware, software, and systems that make modern artificial intelligence possible. This tag covers everything from compute and storage architectures to networking, data pipelines, and observability stacks that keep AI workloads reliable and efficient.
Stories here dig into practical questions: how to design scalable training and inference clusters, choose between GPUs and emerging accelerators, manage feature stores, and orchestrate distributed workloads. You’ll find discussions of MLOps practices, cost optimization, performance tuning, and the trade-offs behind different infrastructure patterns.
Whether you’re building a new AI platform or evolving an existing stack, this tag helps you understand the components, constraints, and design decisions that sit underneath AI products. Reading these pieces will give you concrete examples, architectural patterns, and lessons learned that you can apply to your own systems.
Kiwi AI Infrastructure News
Regional stories with direct local relevance
New Zealand academics call for AI office as fears grow
With deepfakes and other harms already reaching users, academics say a dedicated office would help New Zealand regulate AI and join global talks.
Labour unveils AI plan with tougher rules in New Zealand
Public anxiety over jobs, power bills and harmful content is pushing AI to the centre of New Zealand's election campaign.
Contact & CDC explore 250 MW Stratford data centre
The proposed 250 MW facility could bring fresh jobs and new renewable generation to Taranaki, but it still needs consent, finance and tenants.
New Zealand data centre strategy backs AI expansion
Growing use of cloud and AI is driving demand for local processing, as the sector warns New Zealand risks relying on offshore systems.
Datagrid secures substation kit for Southland AI data centre
Long-lead transformers have been ordered to keep New Zealand's largest AI data centre on track for a late-2027 grid connection.
Companies must adapt to rising AI costs, tech CEO says
Rising token-based charges and compliance risks are forcing firms to monitor AI agents more closely, Fusion5's chief executive says.
Analyst Insights
Research and market analysis connected to AI Infrastructure
InfrOS launches tool to adapt cloud infrastructure for AI
MongoDB launches Atlas Agent Engine for production AI
Gartner forecasts AI strain on services, costs & jobs
LG's 2.6MW cooling unit wins NVIDIA DSX qualification
Cloudera named leader in Asia Pacific sovereign AI
Featured News
Celonis banking head: Banks want proof before scaling agents
Celonis argues frontier AI models are too generic for regulated banking, urging firms to weigh cost, risk and process gains before scaling deployments.
Exclusive: Castrol puts data centre experts behind CORE
The shift towards AI-ready liquid cooling is prompting Castrol to target new-build data centres and costly retrofits across Asia Pacific.
Interview: Cloudera says AI agents reshape zero trust
Enterprises must update identity controls as software agents generate more requests and need tightly time-limited access to sensitive systems.
VAST Data targets AI data sovereignty with DataEnclave
Enterprises and governments could run AI on sensitive data without exposing it, as the technology targets sovereignty rules across APAC.
Interview: Cloudera bets on hybrid AI growth in JAPAC
Banks and large enterprises in Asia-Pacific are rethinking cloud-first AI plans as costs, sovereignty and regulation favour hybrid deployments.
Future of AI is distributed infrastructure built on open source
Rising AI demand is exposing cloud bottlenecks, as firms eye in-house open source systems to cut costs and limit vendor lock-in.
Open source is powering the next generation of frontier AI
Rising inference demand and custom hardware needs are pushing AI infrastructure beyond standard cloud setups, boosting open source tools.
.webp)
Apple's Unified Memory powers Kiraa AI's rapid automation
Australian firms can speed month-end processing from days to an hour by running Kiraa AI on Macs instead of costly cloud servers.
ClickHouse customer base soars on back of AI demand
AI agents are driving a sharp rise in demand for the real-time data platform, which added 1000 customers early in 2026 to top 4000.
Expert Columns
The Next Evolution of Data Centres Isn't Technology. It's People
Building Infrastructure for an AI World in Motion
Fluent is not factual: Why AI needs verified data in APAC
How businesses can avoid widespread harm from cloud outages
AI's data centre overhaul forcing a network rethink
The data center math has changed
Fewer servers, more performance: The new logic of the modern data center
AI is creating a semiconductor boom. But who will keep it running?
Most pharma AI is borrowed, ours is built
Why AI infrastructure planning must happen now
Interviews
Interviews and video coverage from the networkRecent AI Infrastructure News
Enteligent unveils PowerCore stack for AI data centres
The stack could lift AI rack compute density by up to 50% as operators face rising power demands and heat from GPU-heavy workloads.
Google launches llm-d 1.0 for distributed AI inference
The open-source stack aims to cut latency and boost accelerator use for AI workloads that now arrive in bursts and compete with interactive requests.
Google Cloud backs Reflection's Beam for coding agents
Developers could see faster, cheaper agentic coding sessions after Google Cloud added day-one support for Reflection's open-weight Beam model.
Cerebras to bring fast AI inference to Oracle cloud
The addition will give Oracle Cloud Infrastructure customers a faster option for AI agents, with Cerebras inference due in calendar year 2027.
Oracle expands NVIDIA AI cloud with Vera Rubin systems
Customers will soon get broader access to NVIDIA Vera Rubin systems on Oracle Cloud Infrastructure, as the pair scale AI capacity for training and agents.
AMD & Meta expand AI partnership with Helios systems
Meta is boosting the computing base for its frontier models with AMD hardware, while AMD will invest USD $1 billion in the equity offering.
Visko runs live AI video model on Google Cloud TPU
The startup said its Orbis system can keep a live video stream ahead of real time, with each chunk finishing before the previous one ends.
Salience Labs raises USD $125m in Bullhound-led round
Demand for AI data centre networking is drawing major chip investors as Salience Labs plans to expand manufacturing and Silicon Valley operations.
Vertiv & Deloitte issue guide to AI infrastructure
As AI projects move from pilots to production, the guide warns that power, cooling and connectivity limits may decide where workloads can run.
Oracle to offer Google TPUs on OCI Cloud from 2027
Enterprise users will gain another AI chip choice in OCI from 2027, as Oracle becomes the first hyperscale cloud outside Google to host TPUs.
Akuity adds fleet management to Kargo Enterprise beta
The feature aims to ease a growing bottleneck for platform teams releasing changes across hundreds of Kubernetes clusters and edge devices.
Nvidia backs open standards for AI factory builds
The move could let data centre builders use common blueprints for high-power AI systems, easing deployment of racks up to 250kW.
Enteligent launches modular power stack for AI racks
It aims to ease a growing power bottleneck in AI data centres, where operators are struggling to fit more compute into the same rack space.
WWT & DDN launch AI storage design for NVIDIA Vera Rubin
It aims to ease power and planning bottlenecks for AI campuses by matching storage to GPU racks and cutting recomputation delays.
SUSE unveils sovereign AI platform with NVIDIA tie-up
Enterprises will be able to keep AI agents and proprietary data inside their own infrastructure under the new SUSE-NVIDIA sovereign AI stack.
DeepL selects VAST Data for AI translation platform
The deployment should let DeepL's researchers work on larger language models and real-time voice translation with less manual data movement.
Lava flags NVIDIA DCGM flaw exposing AI GPU servers
About 2,100 internet-exposed monitoring hosts left AI GPU telemetry open, putting more than 12,000 GPUs and roughly USD $100 million of hardware at risk.
Dell expands AI Factory to push agentic AI on-site
Adoption has climbed to more than 6,500 customers as businesses shift from AI pilots to on-premise production and weigh rising costs.
VAST & Nvidia launch AI agent sandbox for enterprises
It aims to help enterprises run thousands of short-lived AI agents securely while keeping their state, costs and decisions auditable after shutdown.
Dell expands AI Factory with NVIDIA for enterprise use
More than 6,500 enterprises are now using the platform as Dell adds software, deskside systems and servers to help move AI from pilot to production.
