The newly appointed CEO of thin-client vendor, Wyse Technology, has the task of transforming a small, modestly growing desktop hardware vendor into a major software player. And that's what John Kish says he'll do.
Wyse is ranked number one with 37 per cent of the thin-client market, followed by Neoware Systems and HP, according to IDC. Thin clients essentially display applications that are run and managed on centralised servers, using software from Citrix Systems or Microsoft. But few companies have given up PCs entirely for thin clients.
So how does Kish plan to double Wyse revenue to about $US350 million in two years and make it a $US1 billion company in five?
IDG senior editor, JOHN COX, spoke with him as he was packing for a three-week business trip to the Far East.
You were CEO-in-residence at Garnett & Helfrich Capital, a venture buyout firm that invested $US35 million to take a controlling interest in Wyse and installed you as president and CEO. Why?
John Kish (JK): Wyse has lots of expertise in hardware devices at the edge of the network, which enable network-centric computing. The bet that G&F is making is that this model of network computing is becoming more important as we see a blurring between the enterprise and the consumer markets. We'll take a company that's been defined as a hardware company and begin emphasising new software architectures first for the enterprise space, and then extend these into the consumer space.
And what do you mean by new software architectures?
JK: The real value of thin-client computing, if you talk to large-company CIOs, is not in the [thin-client] device itself.
Their question is, 'How can I build an architecture that lets me decide where (on my network) the processing occurs?'
For example, a customer may choose to have everything run centrally for a highly secure application. But for another application, he may want to run part of the application on the edge device, but not allow any data to reside there.
Implementing this kind of flexible architecture requires the ability to dynamically shift application fragments back and forth between the server and the end device. This is technology that doesn't exist in a commercial form today, but Wyse will be bringing it to market in the next two quarters.
And this is real?
JK: We already have some customers. There are a handful that are already in beta test on this technology.
Why is this flexibility such a big deal to these customers?
JK: They can rationalise cost. You can start making decisions about the right kind of device for a particular kind of user. Traditionally, people tended to buy PCs in lieu of thin clients because they thought, 'You can do so much more with a PC instead of a thin client, so why not just get it, and then add Citrix if you wanted a network-centric model?'
So why not?
JK: But they found that the cost of managing those PCs is excessive. The total cost of ownership is 40 per cent to 60 per cent higher for a PC versus a thin client for any particular application type. Companies want to provide the most optimal solution for their end-users. One solution may be having an active browser on a thin client. But for another, you might want to block the browser and access a centrally managed application. Companies want to be able to make that choice for each user or application type.
What does all this mean for the end-user device?
JK: The Wyse thin clients have addressable, flat memory. The question is how do you use this? Today, you use it to store the presentation subsystem. What we're pushing toward is making this addressable space able to host entire applications or parts of applications and switch those parts or fragments in and out of addressable memory on demand. These will be devices and software that can self-modify based on the particular needs of a user or an application.
What does 'self-modifying' mean?
JK: We usually think of the end device today as having an inherent identity: When it's turned on it 'gathers up' this identity locally, by booting Windows XP or whatever. This identity is fixed with the device: It's always an XP device. But if you remove that assumption so that any time the device is turned on it becomes whatever the network told it to become, in terms of the operating environment and applications, it would be a true utility networking device.
All you really need is a flat, addressable memory space, the kind of memory you can buy at CompUSA, and to be able to move stuff in and out of it. And let the network do all the heavy lifting.
So as long as I have a box with a display screen and addressable memory, I can use it as the equivalent of an iPod or a cell phone or whatever?
JK: Bingo. Or as a cash register, a radio frequency ID reader, an intelligent gas pump, whatever.
Who's going to buy your stuff?
JK: In the enterprise space, we'll provide devices and software for this new model. In the consumer space, we see a whole new economy: The carriers are quite interested in being able to use multipurpose devices to provide applications and services to their subscribers. Most of the carriers subsidise these devices. If those devices can be made to do more, the carrier is in a strong position. So, if we're right, the answer to your question is, 'You will.'