Services / Dedicated Servers
Dedicated Servers in Perth, Australia
Bare-metal dedicated servers from our Dell R620 fleet in Osborne Park. Australian-owned, Australian-located hardware with sub-10ms latency for Perth-metro workloads.
Hosted on Green Racks infrastructure in Osborne Park, WA.
What you get
- The whole physical machine — every core, every byte of RAM, every NVMe channel. No hypervisor neighbours.
- Choice of Ubuntu LTS, Debian stable, Rocky Linux, AlmaLinux, or Windows Server, pre-imaged from our templates.
- Out-of-band iDRAC access for power control, console, and ISO virtual media — useful when SSH is the problem you are trying to fix.
- 10 GbE uplink, dedicated /29 IPv4 + /64 IPv6, plus reverse DNS you control.
- Optional managed layer: OS patching, monitoring, log retention, and incident response by our engineers.
- Daily backup-to-rack option, with retention to match your RPO target.
Server configurations
All builds are full-host Dell PowerEdge R620 systems — Xeon E5-26xx v2 sockets, ECC DDR3, hot-swap NVMe-capable bays. Custom mixes are available on request.
Standard
$ — contact for quote
- 8 physical cores (1× E5-2640 v2)
- 64 GB ECC DDR3
- 1 TB SATA SSD
- 10 GbE uplink, 5 TB/month
- iDRAC + dual PSU
Performance
$ — contact for quote
- 16 physical cores (2× E5-2670 v2)
- 192 GB ECC DDR3
- 2 TB NVMe SSD
- 10 GbE uplink, 10 TB/month
- iDRAC + dual PSU
Memory-optimized
$ — contact for quote
- 24 physical cores (2× E5-2697 v2)
- 768 GB ECC DDR3
- 4 TB NVMe SSD
- 10 GbE uplink, 20 TB/month
- iDRAC + dual PSU
If “bare-metal server” is new terminology, the short version is: you are the only tenant on the physical machine.
How it works
From purchase order to ssh-into-it.
Provisioning
Standard configs ship inside one AWST business day. We image the chosen OS, hand you SSH or RDP credentials and iDRAC access, and confirm reachable IPv4 + IPv6.
Hardware
Dell PowerEdge R620 chassis. Xeon E5-26xx v2 sockets, ECC DDR3, NVMe-capable backplane, dual PSU into independent power feeds.
Network
10 GbE uplink per host into our internal fabric. Multi-homed transit, Telstra carrier handoff via Commscope termination, optional Cloudflare front-end.
Performance baseline
Dedicated hardware removes noisy-neighbour variability, so sustained CPU and disk throughput is stable. That same stability shows up in web-vitals metrics for any front-end you put in front of it.
Who this is for
High-throughput database hosts
Postgres or MySQL primary nodes that need predictable NVMe latency and full CPU access. The Memory-optimized build keeps working sets entirely in RAM.
Self-managed Kubernetes
Bare-metal worker nodes for a self-managed k3s or kubeadm cluster, with consistent per-pod performance and no cloud-vendor lock-in.
Single-tenant WordPress at scale
A high-traffic WordPress that has outgrown shared hosting tiers — same stack as our managed WordPress hosting, but on a dedicated host with no co-tenants.
Frequently asked questions
How quickly can you provision a dedicated server?
Standard configurations from our R620 fleet are usually online within one business day in AWST. Custom builds — non-standard RAM, additional NVMe, or specific OS images — take 2 to 3 business days depending on parts on the shelf.
Can I choose my operating system?
Yes. We ship pre-built templates for Ubuntu LTS, Debian stable, Rocky Linux, AlmaLinux, and Windows Server. Bring-your-own ISO is supported via iDRAC virtual media for less common OS choices.
What hardware do you actually use?
The current fleet is Dell PowerEdge R620 — two-socket Intel Xeon E5 systems with up to 24 physical cores, up to 768 GB ECC RAM, NVMe-capable storage, and dual hot-swap PSUs. Hardware is named in our customer-facing material; we do not rebrand reseller stock.
Do you offer managed dedicated servers?
Yes, as an optional layer on top. The default is unmanaged: you get root or Administrator and we keep the hardware alive. Managed adds OS patching, monitoring, log retention, and incident response from our engineers.
How is this different from a cloud VM?
You get the entire physical machine — all cores, all RAM, all I/O — with no neighbours competing for the host. That makes a noticeable difference for sustained CPU work, NVMe-heavy databases, and any workload where p99 latency matters more than average. The trade-off is that scaling up means physical hardware changes, not an API call.
Can I upgrade my server later?
Yes. RAM, disk, and OS changes are scheduled during a maintenance window. Moving to a different chassis class — for example from a Standard to a Memory-optimized build — is handled as a migration with our engineers, usually completed in a single AWST evening.
Related services: Managed WordPress hosting, Colocation in Perth, AI infrastructure. Or read about the company.