UK How a thin client works
A thin client is a small, low-power computer that depends on a server for most of its work. Applications run on the server, either in a shared session or in a virtual desktop, and the thin client sends keystrokes and mouse moves up and receives the screen image back.
The device runs a light operating system, often a locked-down Linux build or Windows IoT Enterprise, plus the client software for a remote display protocol. The common protocols are Microsoft RDP for Remote Desktop Services, Citrix HDX and VMware Blast or PCoIP for Horizon. The server side is usually virtual desktop infrastructure (VDI), where each user has their own virtual machine, or a remote desktop session host (RDSH), where many users share one server operating system.
Because nothing important lives on the device, a failed thin client is swapped rather than repaired, and software updates happen once on the server rather than on every desk.
Thin client vs thick client
A thick client, also called a fat client, is a full computer that runs its applications and stores its data locally. A standard office PC is a thick client. A zero client goes the other way from a thin client and has no general-purpose operating system at all, only firmware for one display protocol.
| Feature | Thin client | Thick client | Zero client |
|---|---|---|---|
| Where applications run | On the server | On the device | On the server |
| Local operating system | Light, locked-down OS | Full OS | Firmware only |
| Local storage | Small flash drive | Full SSD | None for user data |
| Without the network | Little or no work possible | Keeps working | No work possible |
| Management | Central, image-based | Per device | Central, firmware-based |
| Typical hardware | Fanless mini PC | Desktop, laptop, workstation | Small dedicated box |
The choice is mostly about where the work should happen. Thin clients suit many users running the same server-hosted applications. Thick clients suit work that needs local processing, such as CAD, video editing or AI inference on a machine, or sites where the network cannot be relied on.
Advantages and limits of thin clients
Thin clients trade local capability for central control. The benefits and the costs both come from that.
-
//
Central management. One server image to patch and back up, and new users added from the server rather than at each desk.
-
//
Security. Data stays in the data centre, so a lost or stolen device holds little of value. USB ports can be locked down centrally.
-
//
Long life and low power. Fanless designs with low-power processors often run for years and draw a fraction of a desktop PC’s power.
-
//
Network dependence. Every keystroke and screen update crosses the network. A slow or broken link stops work, so wired Ethernet is normal.
-
//
Limited local performance. Video-heavy or graphics-heavy work needs server GPUs or a more capable client.
Thin clients on the factory floor
Industrial sites use thin clients where many operators need the same screens. SCADA and MES systems often publish their operator screens from a central server, and terminals on the line show them. The same pattern runs in control rooms, warehouses and hospital wards.
Industrial thin clients are usually industrial mini PCs or panel PCs running a light operating system and a remote desktop client. Being fanless, with no moving parts, they survive dust and vibration that would stop an office device. Panel PC versions mount in a cabinet door with a sealed front, so the operator sees the HMI where the machine is.
Where a line has to keep running even if the server link drops, a thick client fits better: a local HMI or edge computer runs the screens itself and syncs with the server when the link returns. Our guide to what embedded computers are covers that local-processing model.
What to check before choosing a thin client
Start with the server side, because it fixes the protocol and the client software the device must run. Then check the display outputs against the screens on site, since control rooms often need two to four monitors from one client. Count the local devices too: barcode scanners, card readers and serial equipment need USB or COM ports that the remote session can redirect.
On the factory floor, add the conditions: operating temperature, dust, vibration and the power available in the cabinet. A device that takes a 12 or 24 V DC input avoids a separate mains adapter inside the enclosure. Finally, plan how images and updates will reach every client, so the whole fleet runs the same tested build.
Quick answers on thin clients
Thin clients: common questions
Do thin clients need internet access?
They need a network, not the internet. A thin client needs a reliable link to its server. On a site with its own servers that is the local network; only cloud-hosted desktops need an internet connection.
What is a thick client?
A full, self-contained computer. A thick client runs applications and stores data on its own hardware, and keeps working without a server. A standard desktop PC is a thick client.
What is a zero client?
A thin client without an operating system. A zero client runs only firmware for a single display protocol, such as PCoIP or Blast. It boots in seconds and has almost nothing to patch, but it only works with that one protocol.
Can a mini PC be used as a thin client?
Yes. Install a light operating system and the remote desktop client, and lock down local storage and USB. Fanless industrial mini PCs are often used this way on factory floors.
Can a thin client run Windows?
Some can. Many thin clients run Windows IoT Enterprise with the remote desktop client, while others run a Linux-based thin OS. The applications themselves still run on the server.
We supply industrial mini PCs and panel PCs for thin-client roles and image them with the operating system and remote desktop client as standard. Tell us your server setup, screens and site conditions, and we will specify the hardware to match.