⚠ This website is a work in progress. Some info may not be correct.
Enterprise · malachitevr.win

Servers & enterprise
hardware.

Custom-built enterprise-grade computers, servers, and rack infrastructure for businesses, labs, and teams that need serious hardware done right.

🔸 Rack Servers 🔸 Workstations 🔸 NAS & Storage 🔸 Networking 🔸 Custom Builds
What I Build

Enterprise offerings.

Every system is specced, sourced, assembled, and tested specifically for your environment. No off-the-shelf compromises.

🖼
Rack Servers
1U, 2U, and 4U rack-mounted servers built for data centres, server rooms, and on-premise infrastructure.
  • Dual-socket CPU configurations available
  • ECC RAM for memory error correction
  • Hot-swap drive bays & redundant PSUs
  • IPMI / iDRAC remote management
  • Custom OS & RAID configuration
💻
Enterprise Workstations
High-performance desktop workstations for engineering, rendering, ML training, and professional creative work.
  • HEDT or server-class CPUs (Threadripper, Xeon)
  • Professional GPU options (NVIDIA RTX / A-series)
  • Large ECC RAM capacity
  • Fast NVMe RAID storage
  • Quiet, thermally optimised chassis
💾
NAS & Storage Servers
Network-attached storage solutions for backups, media libraries, shared team storage, and archival systems.
  • TrueNAS / Unraid / custom OS setup
  • High drive-count chassis (8+ bays)
  • ZFS RAID for data integrity
  • 10GbE networking support
  • Remote access & user management config
🌐
Networking Infrastructure
Managed switches, pfSense / OPNsense routers, VLANs, and structured cabling for office and lab environments.
  • pfSense / OPNsense firewall setup
  • VLAN segmentation & QoS
  • Managed switch configuration
  • VPN (WireGuard / OpenVPN)
  • Wi-Fi 6 access point deployment
🤖
AI & ML Systems
GPU compute nodes and inference servers for machine learning workloads, model training, and local AI deployment.
  • Multi-GPU configurations
  • NVLink & high-bandwidth interconnects
  • CUDA / ROCm environment setup
  • High-speed NVMe scratch storage
  • Liquid cooling for sustained GPU loads
🔒
Virtualisation Hosts
Proxmox, VMware, or Hyper-V hosts for running virtual machines, containers, and hybrid cloud workloads.
  • Proxmox VE / VMware ESXi setup
  • High core-count server CPUs
  • Large RAM pools (512GB+)
  • SAN / NAS shared storage integration
  • Clustering & HA configuration
Component Standards

What goes inside.

Enterprise builds use server-grade components throughout. Every part is sourced from reputable vendors with proper warranties.

Component Enterprise Standard
CPU AMD EPYC / Intel Xeon Scalable or HEDT (Threadripper PRO) depending on workload
Dual-socket options available for rack servers
Memory ECC Registered DDR4 / DDR5 — minimum 64GB, scaling to 1TB+ per node
ECC mandatory for all server and workstation builds
Storage Enterprise NVMe SSDs (Samsung PM9A3, Micron 7450) + SAS / SATA HDDs for bulk
RAID configured via hardware HBA or ZFS software RAID
Networking Dual-port 10GbE or 25GbE NICs as standard; InfiniBand available for HPC
IPMI / BMC dedicated management port on all servers
Power Redundant 80 PLUS Platinum or Titanium PSUs with hot-swap capability
Proper load balancing between PSU pairs
Chassis Supermicro, Rosewill Server, or Fractal Design Node series depending on form factor
Rail kits included for all rack-mount builds
OS & Software Ubuntu Server, Debian, Rocky Linux, Proxmox, TrueNAS, or Windows Server
Fully configured, tested, and documented before delivery
Who It's For

Built for serious work.


01
Small & Medium Businesses
On-premise file servers, backup solutions, and business workstations without the enterprise price tag.
02
Research & Labs
High-core-count compute nodes, GPU clusters, and data storage for scientific and academic workloads.
03
Game Studios & VFX
Render farm nodes, shared NAS storage, and powerful artist workstations for production pipelines.
04
Homelab & Self-Hosting
Virtualisation hosts, NAS builds, and networking gear for serious homelabbers who want proper hardware.
05
AI & ML Teams
On-premise GPU compute for training and inference, avoiding cloud costs for recurring workloads.
06
Colocation
Rack-ready servers built to spec for deployment in third-party data centres, with IPMI for remote management.
Get in Touch

Let's spec out
your system.

Tell me what you're trying to run, how many users or workloads you need to support, and any constraints (rack space, power, budget). I'll come back with a full spec and quote.

Helpful to include: What the system will run (virtualisation, storage, ML, etc.), expected number of users or workloads, rack space or physical constraints, power budget, and timeline.