Stuart Long, Chief Technical Officer, Oracle Systems, Asia Pacific and Japan talks to Dataquest about why, in-memory devices will have a pervasive impact on enterprise computing landscape. Excerpts
During Oracle Open World one heard about the global markets push. Any thoughts on India and your strategy here?
What we're really seeing across the region is basically a change in the way customers are using IT in mobile, social and BI. And that is driving a different type of utilization of IT. Social is unstructured, so you need a way of understanding what people are talking about within unstructured data. And so there is a very big market out there if you're going to do social banking, social retail around building those communities. To do that we need a very different type of system. In other parts of the region we're bringing up 4G - that allows you to do things at much greater speed. We believe that we are moving to a zero latency or real-time type of environment. It gives you a competitive differentiation, because you can start to see what's happening in social media and BI in real time.
In-memory is the in thing. How will it change business practices?
In Open World we announced a lot of in-memory products. What's driving in-memory products is that it reduces latency. We often talk about our performance and our benchmarks. But we need to do this at a very low latency. In the US, the next 3-4 years 40% of the workforce will change. Giving them a web-front end is no longer acceptable. They use devices with interface where you can use pinches, flicks, and ticks. And also these things are GPS-enabled, so you're able to take photos and videos. That changes business practices.
Today, Telcos are under the impression that voice and data are no longer making money so they're going to have to offer differentiated services on that platform to compete in the new market.
How has cloud changed businesses?
Now we're allowing people to do things at an agility they normally don't have access to internally. If I do that project internally then that might take 6-9 months, to build, set up, test, and run. So now we're saving hundreds of thousands of man hours every year from where people would have to do it from scratch. However, cloud can become very expensive. What we're saying is that if you're not using cloud correctly and you're unable to turn off your application, it can become very expensive.
For instance, I run the solution center for Asia Pacific and all my engineers want to demo their product. Now, they want a flexible demo environment, so they go to Amazon and do the right thing by choosing the cheapest of servers, but they don't know that when you turn the server off, your storage went away. Maybe you have to build the thing again. I think the people just kept bleeding inside the track. And that churned a lot of money. What we're doing now is we're proving them an internal private cloud. Initially, if I'm a technologist I want to build this myself.
Any thoughts on cloud adoption dynamics in India?
India is very interesting. When I talk to the large SIs, they host a lot of overseas data here. We're waiting to see if the government changes and says, "hang on if I can let your data come overseas maybe I can let my data go overseas too." So I think there is some regulation change that's required. Most customers are spooked by the security aspect. They're looking at private cloud as some place where they have some control.
So you're saying that Indian businesses are more likely to opt for private cloud?
I think that's the first step. Eventually they will be less worried about the security aspect because cloud is transformational. People will use cloud, to the point where they will have a hybrid kind of approach. Now, transactions go up during festival week by 7 to 9 times. But that means for the rest of the year I have a huge amount of excess capacity. So those models will be worked out. I think any legislation will be positive if the data has to stay here.