Simon’s Blog

If the reports are correct, after its first day of trading, VMWare now has a market cap of around $20B. That’s a rather high value for a company that’s on track to have revenues of only $1B this year. Of course, right now, VMWare’s revenues are growing at a massive rate, and virtualization is a really important sector in Enterprise computing; which is presumably why there is so much interest in the shares. However, I think it’s interesting to look at the fundamentals of the virtualization sector. I’m far from convinced these support such a high valuation for VMWare.

The problem for VMWare is that while, a few years ago, developing virtualization software used to be an incredibly hard technical challenge, today, that’s just not the case. That’s because chip manufacturers Intel and AMD are now building hardware support to make life easier for developers of virtualization systems. Today, VMWare has a lead in this sector. In the not too distant future (perhaps a year or two), you should expect free/low cost/open source competitors to have caught up with where VMWare are now. Xen, Solaris and others are looking pretty interesting. In particular, there’s a whole variety of interesting virtualization work (Solaris-only, Solaris/Linux, and Solaris/Linux/Windows) going on in the Solaris community at the moment. Whether VMWare can move fast enough to maintain a technology lead that’s commercially valuable is a huge question.

Some people are comparing the virtualization sector to the search sector, with VMWare being the “Google”. I think a better analogy is the Java EE app server market. Eight years ago, if you wanted a high-quality Java EE app server, you had to pay $20,000 per CPU for a license. Today, if you want, you can pay zero. I’d be very surprised if you needed to pay more than zero for a high-quality, highly usable, multi-OS virtualization system in two to five years.