Home/Services

Services

Build with us, host with us, or just rent the compute.

Four service families that share one engineering team, one set of facilities and one monthly AUD invoice.

Infrastructure development

From energisable land to a commissioned hall.

AI data centre design & build

Site selection, Tier III+ MEP design, construction management and integrated systems testing for 1 MW modular launches through GW-class campuses, delivered in phased power blocks.

DESIGN · BUILD · IST

Power & grid connection

Transmission and distribution connection enquiries, connection-option studies, system-strength and contingency analysis, terminal-station pathways, connection and use-of-system agreements.

EHV · HV · NER

Energy & microgrid

Behind-the-meter generation, gas firming, grid-forming battery storage, renewable PPAs and take-or-pay supply structures for large-load certainty.

GAS · BESS · PPA

Modular data centres

Transportable 2.5 MW modules with integrated distribution, cooling and security — deployable on regional and behind-the-meter sites, redeployable when the site changes.

2.5 MW MODULES

Liquid cooling engineering

Direct-to-chip and rear-door heat-exchange design, CDU plant, secondary loops, manifolds and leak detection for 80–150 kW racks; immersion on request.

DLC · RDHx · CDU

Commissioning & burn-in

Load-bank IST, then real GPU burn-in — NCCL, HPL, thermal soak — on every node before the first customer job is scheduled.

IST · NCCL · HPL

Data centre services

Colocation, connectivity and hands at GPU density.

High-density colocation

Quarter rack to private suite. Dual A+B feeds, 30 kW air-cooled to 150 kW liquid-cooled racks, redundant cooling, 99.99 % availability design.

GPU capacity hosting

Turn-key hosting for GPU owners, neoclouds and enterprise AI teams — power, cooling, network, security and smart hands in one monthly AUD fee.

Private cages & suites

Dedicated caged space and private suites with biometric access, separate metering and tenant-controlled security zones.

Smart hands 24×7

Australian engineers on site: hardware installation, cable management, power cycling, diagnostics. < 15 min P1, < 4 h general.

Cross-connects & IP transit

Carrier-neutral meet-me room, 40 Gbps+ cross-connects, IP transit on an Australian backbone with access to major domestic and international carriers.

Dark fibre & business fibre

Dark-fibre rings to metro exchanges and major data centres; business and fibre internet for offices and edge sites.

VPC & SD-WAN

Private virtual networking between racks, clouds and sites; managed SD-WAN for multi-site enterprises.

DDoS & firewall

Volumetric DDoS scrubbing and managed next-generation firewall at the edge of your tenancy.

Compute & cloud

Sovereign AI, sovereign cloud, HPC.

Hosted in Australia, billed in AUD, no prompt caching, no training on your data. Reserved or on demand.

GPU compute

Single GPU to full HGX B300 / B200 nodes, hourly or reserved. NVMe-backed local storage.

GPU clusters

Multi-node training pools on InfiniBand XDR or 800G Ethernet, Slurm or Kubernetes, reserved by the month or the year.

GPU-as-a-Service

Reserved GPU pools with AUD billing and a fixed price for the term — no exchange-rate exposure on your compute line.

AI model hosting

100+ frontier and open-weight models — Qwen, DeepSeek, Gemma, Mistral, Kimi, GLM, Phi and more — served from Australian GPUs with an OpenAI-compatible API.

Bare metal dedicated

Single-tenant servers, no hypervisor overhead, full hardware control.

Virtual machines

Dedicated or shared vCPU on enterprise hardware, sub-60-second provisioning, 100 % allocated RAM options.

Private cloud

Dedicated OpenStack tenancy with ISO 27001-aligned controls, hybrid bridge to your colocation racks.

Block & object storage

Ultra-low-latency NVMe block storage and S3-compatible object storage, replicated within Australia.

DR & backup

Offsite, immutable backup and disaster recovery with Australia-wide replication.

HPC

Compute that doesn't queue — CPU and GPU partitions for simulation, research and analytics.

VFX rendering

Burst rendering capacity for studios and post-production, billed by the frame-hour.

Networking & security

VPC, private interconnect to your racks, DDoS and firewall as a service.

Managed & professional services

The people who built it, keeping it running.

GPU fleet management

24×7 NOC, DCGM monitoring, firmware and driver lifecycle, RMA with hot spares, scheduler operations and monthly utilisation reporting — for fleets you own or we host.

99.9 % NODE SLA

Managed services

Managed infrastructure, OS and platform layers, patching, backups and security monitoring for colocation and cloud tenants.

Professional services

Architecture design, migration, cluster sizing, cost modelling and AI workload onboarding by the engineers who run the platform.

Lab-as-a-Service

Test-drive hardware configurations, appliances and reference architectures in our shareholder's lab before committing capital.

IP-appliance design

Productise your software as a hardware appliance — design, validation, manufacturing and delivery through the Hyperscalers channel.

24×7 support

Australian-based engineers, one phone number, one ticket queue across colocation, fleet and cloud.

Industries

Who we build and run compute for.

Sovereign workloads

Government, defence-adjacent and regulated data that must stay in Australia under Australian control.

AI startups & neoclouds

Reserved GPU capacity and hosting for teams that need hardware faster than they can build a hall.

GPU asset owners & lenders

Hosting, fleet management and utilisation reporting for financed GPU fleets.

Financial services

Low-latency, ISO 27001-aligned compute and private cloud with AUD pricing.

Enterprise

Private AI, private cloud and colocation with a hybrid bridge to existing estates.

Software & MSPs

Wholesale compute, VPC and white-label infrastructure for partners.

Research & education

HPC and GPU partitions for universities and research institutes.

Media & VFX

Burst rendering and high-throughput storage for studios.

Not sure which service you need?

Describe the workload and the timeline; we'll map it to the chain and quote the pieces you don't already have.

Talk to engineering