Thanks for visiting Scientific Computing World.

You're trying to access an editorial feature that is only available to logged in, registered users of Scientific Computing World. Registering is completely free, so why not sign up with us?

By registering, as well as being able to browse all content on the site without further interruption, you'll also have the option to receive our magazine (multiple times a year) and our email newsletters.

Tools to make things simple

Share this on social media:

Brian Ritchie, President and CEO of Blackstone Computing, tells Tom Wilkie how to get scientists doing more science and less computing.

Are clusters, the Grid, and peer to peer, supplanting supercomputers?

We're having executive-level meetings with customers about strategic trends and directions. P2P, the Grid, and distributed computing are being used interchangeably. Customers love the concept but are not clear about how to get from here to there. Before you can go to wider collaboration you have to be sure that your own IT is up and running and solving primary problems. Life science companies don't have their house in order, in terms of their own core IT, for them to leverage the external resources. There are very few companies that have their IT infrastructure in place and at high efficiency.

 

The market is very fragmented and the life sciences are at an unusual stage. The majority of tools available are nowhere near as developed as in other areas and customers are doing a lot of development themselves. Customers need to bring 'production computing' discipline to the life sciences. They need to be able to accelerate discovery and can only do that by having strong repetitive processes.

 

Scientists spend too much time understanding 'how do I access data from other databases' and so on - computer-science oriented activities. A strong IT infrastructure takes the informaticist out of computer science and back to doing more science.

 

That's the core problem. How do you access, massage, use, that data with the right applications on the right hardware? Users are trying to have the right applications on the right topologies - client server; compute farm; the Grid.

 

Will researchers take kindly to 'production computing'?

 

Customers will try to run applications in a non-traditional manner: applications that can be parallelised or spread across more computers. Then they'll say they want to do more - more computers; more users. This is the production phase: 'Let me use this across the whole environment'; then comes: 'What other applications can I utilise?' That's when they start to look at enterprise-phase P2P and distributed computing, and then out to a broader Grid-type environment.

 

How does Blackstone fit in?

 

Customers have to go to lots of different places to put together an IT infrastructure. Customers are trying to work out how to fit this together. We provide the consulting 'glue' to stick these together but now we can also provide 'product glue' - we have taken a lot of our consulting skill and codified that into software products that we can leave behind.

 

PowerCloud is a suite of applications that allows you to plug in the data analysis tools you want; the hardware topology you want; so your scientists don't have to worry about that. We've provided that kind of value from a consulting standpoint for years. But once we got the IT environment tuned, what happens when we leave? Scientific objectives change, because you discover things. We say it's great for us that they keep inviting us back. But now we've built something, PowerCloud, to enable you to plug into your infrastructure whatever you want: pipeline tools; data access tools; computing middleware; and different hardware topologies. We've developed applications to allow customers to do what we used to do as consultants.

 

Will you put yourself out of business?

 

We feel that the market need is greater than we can cover through consulting. We've been evolving our company from a purely consulting to a product software business with supporting services. There's a lot of problems to be solved. Collaborations between vendors and suppliers are needed; and emerging standards to get to the ideal. Globus is already doing some stuff.

The choice is vendor-independence or monolithic suppliers. Some players say: 'This stuff is complicated, so you have to lock in to us.' We say: 'You need tools to make it simple, so you can do it yourself.'