Thin is in


By Marc Doehnert, ANZ Business Development Manager
Monday, 12 January, 2015


With the decline in sales of personal computers and the rapid rise of cloud computing, many businesses are turning to thin clients. So what does this mean in terms of connectivity, cabling and power? Read on to find out.

Thin clients rely on a server to provide most of their computational processing, unlike a traditional PC or ‘fat client’ that processes information by itself. Therefore, all thin clients need to be connected directly to a ‘host’ computer, usually in the form of a server, a group of servers or an outsourced cloud storage facility. This requires connection from each thin client to the network.

By definition, traditional network architectures will have to change as more and more organisations move towards a virtual desktop environment.

Devices can now move seamlessly from one access point to another without interruption to the session; so, for example, a thin device operating on a forklift can move seamlessly throughout the warehouse, picking up connectivity from each WAP that it encounters.

Migrating to a thin client computing environment can significantly reduce a company’s need for bandwidth. However, depending on how the network is set up, it can move the demand on the LAN/WAN infrastructure to the company’s intranet infrastructure.

Thin clients reduce the cost of ownership in organisations where traditional computers have more power and functionality than the infrastructure requires. With limited or no onboard storage, thin clients cost less to buy and less to run, and they tend to be more reliable, allowing companies to capitalise on their consolidated storage environment. Security is also simplified, since it now takes place at the server level rather than on each individual device.

A Microsoft study conducted by NEC and Groupe Bull shows that the highest user of bandwidth in a company is a structured task worker, typically performing the same tasks repetitively such as claims processing, handling accounts or customer service. These applications are typical of the type of deployments where a thin client network performs very well for an organisation.

Printing can also spike the amount of bandwidth that a staff member typically uses, up to as much as 200 Kbps, and other network demands such as multimedia and VoIP need to be factored in as well. The complexity of having multiple printer drivers in a virtual solution such as Citrix XenApp can also have a fairly severe impact on the stability of a network. Especially in regional offices where redundant bandwidth is at a premium, printing a document can cause a big spike in the use of the network link, which can impact the experience of other users on the network. Again, the message that needs to go out to end users is clear - carefully consider how much bandwidth the thin-client network is going to need. When allocating bandwidth across a network of thin clients, a fairly universal figure is to budget 100 Kbps for each end user, plus extra for any other daily functions that will place additional demand on the network.

So migrating to thin-client computing (TCC) does not necessarily reduce the need for a sophisticated communications infrastructure. To set up an effective thin-client environment, an organisation will have to invest in or already have servers, routers and high-speed connections, since the user experience will be greatly affected by the state of the network and issues such as latency and packet loss.

From a company’s point of view, cost savings will take place in the reduced cost of the hardware itself and the reduced power costs associated with thin clients. Factoring in the additional server infrastructure required to run a thin-client network, on average each device consumes between eight and 14 watts. This compares to a fat client or desktop PC consuming between 60 and 250 watts. The electrical system load is therefore significantly reduced for an environment running thin clients and can be planned and configured accordingly. In cases where a new office is being planned or an existing premises expanded, the number of power points can potentially be reduced by consolidating end-points, since the power draw on any terminal is going to be substantially less than if it were running a fat client.

Moreover, the fact that thin clients can be remotely administered over a WAN frees up time for IT administrators to focus on other aspects of the network and IT infrastructure.

Another physical benefit of the thin client is the potential for them to operate in harsh environments. Since they have fewer moving parts and tend to be more robust, there is less that can go wrong, so exposing them to hazardous conditions causes fewer risks to business continuity. However, the servers that host applications for the thin-client network will still have to be in a safe, shielded environment, which is another point of consideration when planning data and electrical networks.

Since thin clients cannot operate without connectivity to the servers, an uninterrupted power supply to the network and server room is of vital importance to a business. In remote areas and cases where power is unreliable, connecting to an off-site cloud might be a good option since it removes one risk factor from the equation.

If connectivity between the client and the server is broken, clients cannot continue working, although the current session will remain in the state it was in at the time the link was lost. The same goes for fat clients, of course, in cases where applications are centralised - ERP software, hotel reservations for example.

Therefore, LAN/WAN resilience is vital. In high-availability environments, best-practice recommends alternate path and supplier routing, particularly for WAN links. This may increase initial costs for deployment over a fat-client environment, as it relies on each link being initiated. However, since each network link using a centralised service such as VDI or presentation virtualisation does not need as much bandwidth as traditional PCs, ongoing costs will again generally be lower.

The fact that thin clients are practically plug-and-play makes for very easy installation, so upsizing a network is fairly straightforward. The thin clients provided at IGEL are automatically configured as soon as they are attached to the network, so expanding a premise is a matter of providing a power source and enough bandwidth, then simply connecting the new devices to the network. There are many benefits for companies planning a move to thin clients or a virtual desktop environment, but careful planning of network requirements is vital.

Related Articles

Meeting demand for off-grid power generation

With power generators in high demand for civil engineering, an Australian dry hire company needed...

Protecting wildlife from electrical assets — and vice versa

While overhead lines are cost-effective and less carbon-intense than underground cable, the...

Why power networks need a 'smoke alarm' system

RMIT University has designed an early fault detection system for powerlines that has been...


  • All content Copyright © 2024 Westwick-Farrow Pty Ltd