Will local assemblers go for Lintel over Wintel?

Market pundits have been pushing the benefits of Linux for a long time, but although there has been a lot of noise about the open source operating system coming from the US and some of the big name vendors, it has been difficult to see Linux as a credible alternative to commercial Unix for companies in the Middle East.

  • E-Mail
By  Mark Sutton Published  July 30, 2003

Market pundits have been pushing the benefits of Linux for a long time, but although there has been a lot of noise about the open source operating system coming from the US and some of the big name vendors, it has been difficult to see Linux as a credible alternative to commercial Unix for companies in the Middle East.

Apart from some interest from universities, Linux has not attracted the same sort of high profile support as it has elsewhere. But, quite quietly, and without much fuss, this has changed.

For example, the Government of Bahrain has deployed Linux for its systems. Governments in other parts of the world have been among the vanguard of Linux adoption, so the government of Bahrain is not sailing into uncharted waters—open source as an alternative to commercial operating systems has gained plenty of supporters in countries where government spending is scrutinised.

Once the first few reference sites are in place, it is easy for others to follow on, but even still, it marks a major show of support for an end user of the size of a government to pick Linux.

Perhaps more surprising is the decision of Standard Chartered Bank to run its retail banking system on Linux. When a major multinational bank decides to run a core application on Linux, then you know that things are changing.

The vendors are also getting behind Linux in a much more practical way than before. Oracle is promoting Linux on its performance, not just its lower TCO. IBM is investing heavily in universities around the region to help create the beginnings of a Linux skills pool for the Middle East.

So where is the channel in all this? There are a handful of systems integrators that are working with Linux, although so far, demand has not quite created a return on investment for channel partners. With the backing of so many vendors and with commercial usage of Linux on the rise however, this looks set to change.

Perhaps even more interesting is the potential of Linux at the lower end of the market. Systems builders are turning to server assembly in large numbers—when a white box Intel-based server can cost as much as 30% less than an equivalent brand name server, it is not surprising.

Local assemblers can, with a bit of investment in hardware skills, move into server production, to make much better margins for themselves, while providing even better prices to their customers. The question is, whether they can go one step further and put Linux on the servers they build? Given the cost saving of Linux over Windows, it certainly seems attractive.

Of course, white box assembly is almost totally focused on Wintel architecture, but until very recently local assembly was only focused on producing desktop PCs, not servers. Intel is doing a good job of giving local assemblers the expertise to produce the hardware, but as far as the operating systems go, there is no one to champion Linux. For all of IBM’s dedication to Linux, it is difficult to see why a hardware vendor would back local assemblers.

This leaves the task in the hands of the commercial Linux vendors such as Red Hat, and the assemblers themselves, and this is where I suspect the concept of local assembled Lintel servers falls flat. For all of the enthusiasm that local assemblers are showing for server building, they haven’t moved of their own accord—it has taken a marketing shove and promotion of the business benefits from Intel to get the ball rolling.

Given that Intel already had the ear of the local assemblers, and its marketing dollars, the Linux vendors have got their work cut out to convince assemblers to switch to a new operating system. They are also going to have to make Linux administration skills as prevalent as Windows skills, no easy task either. For the time being, local assemblers have no great impetus to change, and no skills to draw on if they wanted to change. Local Lintel could be a way off yet.

Add a Comment

Your display name This field is mandatory

Your e-mail address This field is mandatory (Your e-mail address won't be published)

Security code