HP: Why Converged Infrastructure Is Better

Moving to a converged infrastructure delivers significant savings, but HP still believes in open standards, according to Mark Potter

Customers want flexibility

While touting the benefits of converged infrastructure, Potter was keen to point out that HP does not lock its customers in to a single architecture.

“[HP] BladeSystem is a good example of converged infrastructure – we have over 300 partners. So Blade network technologies has products that fit in here, Brocade, Emulex, QLogic, Microsoft, VMWare…” he said. “So just because you design something to be converged it doesn’t mean that you prevent others from participating in that ecosystem. That’s important because customers do want that flexibility.”

Despite this, however, HP does provide a strong incentive for customers to move entirely over to its own converged technology. “Lets say that you were wanting to buy your networking and the edge switches separate from your server separate from your storage, you wouldn’t get the savings,” said Potter.

“When you take a step back and you have the expertise like HP to design it all together, you get tremendous savings. Because if you think about a server connected into the network, all it’s doing is using very expensive cabling and PHY technology to take high-speed digital signals, convert it to analogue, pass it over a copper cable or a fibre cable to another very expensive transceiver that converts it back to digital. And then it gets processed and sent somewhere else.

“So when you converge it you get rid of all that, plus the management, plus all the firmware that you have to take care of, the lack of ability to change, the failure scenarios… they all go away, because you design the servers to connect directly into the fabrics and aggregate right there. So you can do something with two parts, and dramatic savings on power and cooling – versus the old way of doing it because I want to buy everything separate, because I feel like a need that “choice”, and you get 217 parts.”

According to HP, customers are currently spending approximately $10 billion on maintenance and operations, and that is expected to grow another $1 billion over the next two years to 2012. By reducing sprawl and improving efficiency, HP claims it can free up a substantial slice of that money for innovation. “Many of them look and say, why would they not do this?”

Technology for the cloud

BladeSystem Matrix software provides a very simple user interface that enables users to drag and drop pieces of the infrastructure in order to build an architecture, according to Potter. It is then able to automatically configure all the servers, connect them to the right networks and storage technologies, and then apply applications on top.

This is particularly useful for cloud vendors, who need to be able to connect their private cloud to a public cloud. “Imagine if you’re a customer and you’re running a private cloud and you need a little bit more resource, wouldn’t it be nice if you could take that same template you’ve created and point it somewhere else and they do it on your behalf. Same network, same servers, same storage, that’s configured to your standard – but now you’re able to do it anywhere,” said Potter.

When asked where he thought the industry was headed with regard to converged infrastructure, Potter remained tactful.

“Where the industry is going plays very very favourable to where HP is investing and what customers see us as, which is the open industry standard server, storage and network provider,” he said. “I think we’ll continue to build open industry standard platforms like we do today, they will be designed from the beginning to optimise around converged infrastructure. But what HP’s known for is open industry standards and giving customers choice, and that’s what we’ll continue to do.”