...how much computing power it has, what its network configuration is, and so on, and delivers that configuration to our software, which maintains it in a database.
Then we use those profiles to configure a variety of applications from the application images. Those profiles can be designed to have the right capacity, connectivity and storage to deliver the right performance for the application.
And you can work through that system?
We have developed our software to be integrated with [Cisco's] object manager and we have been using the UCS to do it. It is quite an impressive invention. The great thing is that it relies on a blade architecture, so it supports Linux and Windows and therefore thousands of applications.
It is also very manageable, unlike so many hardware inventions.
How long have you been working with Cisco?
We have been collaborating since early summer of last year. One of the designers of the architecture approached us and started taking us through it and we were quickly convinced.
Is Cisco delivering the blade now?
I think the computer actually delivers next month, April.
So you think the efficiency claims Cisco is making are valid?
It really does take a tremendous amount of work out of the whole thing. One of the brilliant things Cisco did was not to try and be everything to everybody. It found partners. It didn't try and design its own virtualisation or management software, unlike IBM and HP.
The way the market is going, you will have two $100m (£70m) companies, IBM and HP, battling it out, and then Cisco and the rest of us saying open systems are the better way. From our point of view we are in an enviable position. We are providing the only management software at the outset, but it is not exclusive.
What is the deal between Cisco and BMC?
Cisco is reselling our software, which will remain BMC branded. We are committed to doing joint sales to begin with and will be focused on some of the largest accounts in the world.
What about the timing of the announcement, as it is the middle of a recession?
This is happening at a time of highly dynamic virtual environments. HP has called it a time of the adaptive infrastructure. What Cisco said was, with a standard blade architecture, with the right partners, you can adapt.
Cisco said the bottlenecks that inhibit your computer from working properly are in the memory and the network and storage interfaces. So they have put the design around those points in the architecture to give them much higher bandwidth and much better performance.
For example, memory is a bottleneck on most computers. Cisco took a new IC and tripled the memory availability and that has a dramatic effect on performance. What Cisco has to offer will mean a step change in the labour required to run a datacentre.
One of the advantages HP and IBM have is they can both call on huge networks of partners. How can Cisco compete with that?
It is not a problem. Cisco has one of the largest partner networks in the world.
But they are partners in a different way?
Sure, but the important thing to remember is that it is a business relationship and Cisco has the experience in datacentres. It's not like they are going to be selling computers off the shelf.
Those partners will look at this as an opportunity because it will take a lot of their labour out. The labour required to manage all those wires is not value-added labour.
