How to
Practical guides for local AI.
Set up, size and connect AI hardware: written by the people who ship it.
Latest guide
Networking a small AI lab with 25G and 100G
How to connect workstations, GB10 systems and storage at 25G or 100G without a data-centre budget: which switch, DAC vs optics, and how breakout cables work.
-
RTX PRO 6000 Workstation vs Max-Q: choosing a GPU for local AI
Same chip, same 96 GB of memory, different power and cooling. When to choose the 600 W Workstation Edition, when the 300 W Max-Q, and how the smaller RTX PRO cards fit…
-
How much memory do you need to run an LLM locally?
A simple rule of thumb to work out whether a model fits on your hardware, what quantisation does, and why context length matters.
-
Which GB10 system should you buy? DGX Spark vs Asus vs HP vs Lenovo vs MSI
Five desktops, one chip. What is actually the same, what differs (storage, operating system, cooling, support), and how to pick the right one for you.
-
How to connect two GB10 systems into a cluster
Link two DGX Spark or other GB10 systems over their ConnectX-7 ports to run models of around 405 billion parameters. The cable you need, the network setup, and how to check it…
-
Getting started with your DGX Spark or GB10 desktop
From unboxing to your first local model: what to plug in first, how the first-boot setup works, how to reach the system from your laptop, and three ways to start running models.
- Need a hand? Ask us anything about your build. Contact us
