Uses

this is a complete inventory of the hardware that powers my homelab, development work, and daily computing.

no marketing, no affiliate links - just the actual machines i use and why they exist in my setup.

the hardware is organized into:

• personal notebook - my daily driver
• home lab main server - the powerhouse for databases and services
• home lab arm64 node - raspberry pi for arm testing
• cloud x86-64 machines - dev and prod in the cloud
• cloud arm64 machines - arm-based cloud workloads

---

Personal hardware

Main notebook

This is my daily driver. I use it for everything related to study, code, SSH, docs and general work.

Specs

• CPU
  • AMD Ryzen 7 7445HS
• RAM
  • 64 GB DDR5
• Storage
  • 256 GB SSD
  • 360 GB SSD
  • used for system, dev tools, local databases, containers and files
• GPU
  • Nvidia RTX 4070 (notebook)

Main use cases:

• terminals and IDEs to work on code and documentation
• SSH into all lab and cloud machines through Tailscale
• database clients, admin tools and dashboards
• browsers and apps that talk to services running in the lab
• some local containers and light VMs when needed

This notebook is the place where I drive everything else.

---

Home lab hardware

Main server at home

This is the big machine in the homelab, focused on databases, services and experiments.

CPU

• Intel Xeon W9 3475X

Memory

• 256 GB ECC R
• 256 GB ECC R in mirror configuration
• total of 512 GB of ECC memory, configured for redundancy to avoid failures

Storage

Rotational storage:

• 14 TB HDD main data
• 14 TB HDD for redundancy
• 14 TB HDD for backups

NVMe storage:

• 256 GB NVMe for the system
• 500 GB NVMe used as file cache
• 4 TB NVMe for cold files and long term storage

GPUs

• Nvidia GT210 for video output
• Nvidia GT 1050 for small workloads and video codecs
• AMD Radeon RX 7900 XTX for AI workloads

note: planning to replace all nvidia and intel hardware with amd components in the future

typical usage:

• PostgreSQL and other databases in real lab style setups
• Docker/Containers stacks for self hosted services
• monitoring, logging and backup jobs
• AI experiments using the RX 7900 XTX
• storage for media, backups and lab data with redundancy

This is the core of the home lab and where most heavy work happens.

---

ARM64 node in the home lab

I also keep an ARM64 environment at home for testing and variety.

Board

• Raspberry Pi 5, 16 GB version

SoC

• Broadcom BCM2712
• quad core ARM Cortex A76

Memory

• 16 GB LPDDR4X 4267

Storage

• 64 GB eMMC module for the system
• 512 GB drive for containers and files

Use cases:

• running ARM containers
• testing services on ARM64 before going to cloud ARM
• small always on services with low power usage
• experiments with storage on small boards

This node helps me make sure my setups are not x86 only.

---

Cloud hardware

I also run dev and prod workloads in the cloud. Right now there are two main groups of VMs.

x86 64 dev and prod in the cloud

These are two virtual machines with the same hardware, one for dev and one for prod.

Specs per VM

• CPU
  • AMD EPYC 7551, 8 vCPU
• RAM
  • 16 GB
• Storage
  • 250 GB for system, data and logs

There are two of these machines with the same configuration.

Main use:

• running databases and services in a more realistic production like environment
• testing deployments that talk to the home lab through Tailscale
• simulating scenarios where part of the stack is on premises and part is in the cloud

---

ARM64 dev and prod in the cloud

I also keep ARM64 machines in the cloud for tests and workloads that make sense on that architecture.

Specs per VM

• CPU
  • 8 vCPU
• RAM
  • 24 GB
• Storage
  • 150 GB for system and data

note: exact cpu model not specified by cloud provider - what matters is the performance and arm64 architecture

used for:

• running ARM64 containers and services
• checking performance differences between ARM and x86 setups
• experiments with energy efficient style workloads

Again, the idea is to have at least dev and prod separated in this group.

---

Why this mix of hardware

The hardware is not random. It follows a simple idea.

• the notebook is the control center
• the home lab main server is where heavy work and storage live
• the home ARM64 node gives me variety and low power experiments
• the cloud machines let me simulate more realistic production setups

With this combination I can:

• test databases and services with serious memory and storage at home
• run AI workloads locally using the RX 7900 XTX
• keep redundancy and backups with mirrored RAM and multiple HDDs
• compare x86 and ARM64 behavior
• practice dev, staging and prod style separation

---

this inventory helps me remember what i actually have and serves as documentation for my infrastructure. if you are reading this and you are into homelabs, now you know what is running behind the scenes at yuricunha.com.