This is a transcription of an interview of Prith Banerjee, Hewlett-Packard Senior VP and Director of HP Labs, by SiliconAngle founder John Furrier that was webcast live from the HP Discover 2011 conference on SiliconAngle.TV. Banerjee discusses the reorganization he brought to the Lbs four years ago, what the recent change in leadership at HP has meant to his organization, his vision of the future, and some of the things that HP Labs is working on today that may become tomorrow’s leading-edge products.

JF: Tell us where we are now with HP Labs. You did a reorganization four years ago around organizing your teams, so give us an update on that, and then we’ll talk about the future.

PB: HP Labs is the corporate research arm for Hewlett-Packard. I’ve been associated with HP Labs for about 4 years. During that time it has been through an interesting strategic transformation. As you mentioned, four years ago we took all the creative people at HP Labs and focused them on about 20 high-impact research projects around 8 teams. The teams are:

1. Print and content delivery, where the future is moving from the analog world of printing to the digital world of printing.
2. Mobility and the future of mobile devices and how technology will change in the next several years.
3. Cloud and security, which are becoming very important today.
4. Information analytics including structured & unstructured data and big data, another area that is becoming very relevant today.
5. Intelligent infrastructure, where we are going to collect a lot of information around nanoscale infrastructures – collect the data, store the data, process the data, and essentially create a very intelligent infrastructure for our customers in the future. That seems like a good call also.
6. Networking and communications, the core technologies for bringing data from mobile devices to the cloud and so on.
7. Services focused on how you can use technology to create much better services in the future around certain verticals.
8. Sustainability, which we felt was so important to society: lowering the carbon footprint, lowering consumption.

So we have been at it now for about three & a half years and made substantial progress. And the good thing, as you know, Leo announced strategy for the company around seamless, secure computing for the connected world. It’s around pulling in cloud computing, which is the enabler for computing, storage and so-on for all the billions of mobile devices out there, and what you are going to do with it – analyzing information.

So you look at our eight teams, they are so well aligned with HP’s strategy. Of course as the forward-looking part of HP that’s our job, but we made some good calls four years ago.

New Leadership

JF: I have to say I met with Prith when SiliconAngle was just me, and we had a tour and got a chance to talk with some people, and you really made a good call. Not only does it match up with HP’s strategy, it’s really lining up with the megatends that are lifting us out of this recession and into a massive IT and global economic recovery. So it’s super exciting, you have everything covered, it’s all good calls. So today you have a new CEO who is an innovative guy, he likes tech. During the keynote he said “tech is cool” about three times. So you must really like that.

PB: He [Leo Apotheker] has been so supportive of innovation and HP Labs. In fact the day he joined HP he came over to HP Labs. He talked to me, and talked to our fellows and talked about technology, and he has been extremely supportive of HP Labs. And in fact in the Summit on March 14 in his introductory paragraph he talked about innovations coming out of HP Labs. He named the StoreOnce deduplication technology that came out of HP Labs. He named the sensor technology, our nanoscale sensors trying to help one of our customers in the oil exploration area. He talked about the work on memristors. These are innovations that we started at HP Labs and that now are in the process of being transferred to our businesses and the hands of customers, and here was Leo at the Summit talking about them in his opening paragraph. I felt really good about it.

JF: And how does that energize the team? Do they say he gets it? Are they really energized?

PB: They are very excited about him. In fact he is coming to the Labs again for a deep-dive visit. He has been very supportive of all the research. The lab people are very excited. We can’t wait to show things to Leo. I was just talking to ???. He had a chance to talk to Leo about some of his ideas, and he [Apotheker] just said, “Go and do it.” So he’s extremely excited about innovation.

Vertica

JF: I’m very excited for you. Let’s talk about R&D and labs in general. In a lot of companies R&D sort of sits out on the fence with some crazy ideas for the future. They do some good research, but it doesn’t always make it back into the business. That’s changed under your leadership. I’ve heard a rumor that based on the big data killer you had a couple of years ago that you were pretty peaked on some of your research, and then Leo comes in with some other folks and says, “I want to get Vertica.” He buys Vertica, pops it out of the market, takes that of the table, and now I hear rumors that they’re working with some labs teams, some analytics lab teams? Is that true, can you confirm that?

PB: Yes. We are very excited about it. Let me talk about the Information Analytics Agenda from HP Labs, and then I will tell you about how we are thinking about this with Vertica.

So the problem that we identified is about big, fast, total data. Anybody can analyze a gigabyte of data. If you do 1,000 Gbytes, that’s a Tbyte of data. You take 1,000 Tbytes, that’s a petabyte of data. 1,000 petabytes is a zenobyte. So you are talking lots and lots of data, and can you analyze it in real time, as it comes in. You are trying to analyze all this stuff at the speed your business is running, so at the speed of business.

So people in the past have done business intelligence—trying to do statistical correlations between all customers who bought Bounty paper towels and also bought Charmin bathroom tissue. At the end of the week, the end of the month, the end of the year they do this analysis. That’s not relevant any more. You need to do it at the speed of business.

The third thing is around total data, which is structured and unstructured data. If you look at the amount of data being generated today, doubling every 18 months, and 80% of the data is actually unstructured data. So you look at this big, fast, total data coming at you from a variety of sources – the billions of nanoscale sensors, the tweets that are happening on Twitter, the social media, Facebook and so on – how are we going to analyze it all?

So within that we have three big projects:

1. One on taming the information explosion, trying to extract metadata from the data, essentially looking at unstructured data and trying to bring some structure to it and putting it in the structured forms of data.
2. A project on live business intelligence, looking at the amount of data that’s coming in very fast and analyzing that at the speed of business, applying it to domains like what we call live operational intelligence, live customer intelligence.
3. The third one is around IT informatics, looking at the large amounts of data that the IT industry creates -- server going down, network doing something, etc., -- doing data mining on it, and helping the CIO make better decisions.

So this was the research agenda for Labs. We were looking at if you try to build a thing like this you need a high-level integration layer, a layer that will talk to the customers, you need an analytics layer, you need a storage layer, and servers and architecture to build it on. When we were looking at the kind of storage layer that you need, we were following up on the work that is going on in columner stores, and data stores, and in-memory data, and so on.

We looked at Vertica, obviously liked the company, it made a lot of business sense for us to buy it. Vertica’s now actually been integrated and is under Shane Robinson. I report to Shane. Chris Lynch from Vertica reports to Shane. So Chris & I had a good conversation. The engineering teams have gotten together, and we are trying to align our ducks.

JF: So basically you confirmed that there is an initiative going on at Labs. Also Vertica’s too important to give away right now to what makes sense to a division. Is there a bigger picture there? So you have this arc that is Vertica. You can do things with it now – we heard from ESSN has an appliance. Paul Miller talked about that, Dave Donetelli talked about that. They instantly grabbed the big data and shipped the product in like two months. What is the big story with Vertica? That’s a big asset for HP. What’s the vision?

PB: That’s a business decision. So what I can tell you about is how excited we are from a research perspective to work together with the Vertica team and bring its outstanding technology into our development. Our business leaders will then figure out the business models to bring it to market.

Cloud Architecture

JF: Operating systems. Every time a new trend comes in it’s the death of something – the death of the mainframe, the death of client/server, the death of the PC. But it really never changes. Nothing ever dies, it just morphs into a new form. Cloud and big data – cloud, and mobile and social – those big megatrends. What is the future in those areas, because they’re changing. Cloud is now going mainstream – Steve Jobs announced iCloud – that’s going to take the nomenclature to a whole other level. So is there a cloud operating system? We talked about the data operating system in the past with some of your folks. How’s the systems architecture going to change in the future? Is it going to be distributed? Is it already distributed?

PB: My team at HP Labs is working on this exact topic. The Cloud and Security Lab is working on some very innovative IP [intellectual property]. People are now convinced that cloud is a very interesting business model, and instead of owning IT assets, you should pay for it as you go through compute-as-a-service, storage-as-a-service, and so on. We’ve seen some initial versions of public cloud offerings from some vendors, but we at HP Labs have recognized that the real opportunity’s in the enterprise in terms of the amount of data that’s out there. You need to deliver it in a very secure manner, highly scalable matter, highly scalable manner, very reliable, and highly available manner.

So at Labs we are building this enterprise-grade platform. Our internal name for it is “Sirius”. Essentially using virtualization technology, we allow application developers to create virtual compute cells, storage cells. And by having network virtualization, we can enable customers to have these multi-talented applications so that the applications are isolated from each other. That’s extremely important in the cloud. You don’t want my application in the cloud impacting your application on the shared infrastructure. Along with it is programming environments and so on.

There is a desire for enterprises to use the hybrid cloud, which is private cloud for most computations and storage, but periodically when you need to support flexible needs you can go over to the public cloud. How do you do that in a very seamless manner? And those are the technologies that are being worked on at Labs.

Security’s a key enabler. Every customer says they are scared about security.

End-User Computing

Then you talk about how people access the cloud. Essentially it’s through these end terminals, these connected devices – tens of millions of smart phones and tablets and PCs and notebooks and printers and so on, and HP’s bet is around webOS. The same seamless experience on these multiple touch-points accessing data, computations, etc., on the cloud.

The real deal is that people have these two lives. We have the personal life, we do things on the computer as a consumer. We are playing games – Mafia Wars or Angry Birds – on a smart phone or tablet. And you would also like that same device to do things in the enterprise. Now typically CIOs in the enterprise do not want to allow these devices in the enterprise network. So in the enterprise you have to use a PC, for your other stuff you use a smartphone and tablet. We think it should be the same device.

JF: The consumer wants it to be the same device, too.

PB: So at Labs, for example, we have an effort called Managed Cloud Communities where we enable CIOs to manage your same devices, so you as a consumer can use your device for whatever consumer applications you want, and the CIO can enable you to do the enterprise applications, and the two worlds will not intersect. That kind of isolation is so key. Those are the kinds of key applications we re developing at Labs.

JF: And the complexity has to be completely extracted away. It cannot be user interaction, it has to be completely under the covers.

PB: And the future of this technology is around how humans interact with this technology. The reason these smartphones and tablets have become so popular – on cell phones a few years ago you had to use a keyboard to work with data. Now you use touch, and you can expand, delete, move things around in operating systems like WebOS and others. But we at Labs are thinking about the future of interfaces. So we have a project on rich, intuitive interfaces, combining gestures and speech, essentially making it so much more intuitive that it’s like two humans talking together.

JF: We could put a Cube in HP Labs broadcasting all day long, 24/7. Mobility. Obviously you guys saw this coming, you have an initiative in place. Is there anything on the mobility side that has surprised you or has changed your agenda?

PB: When we picked our research agenda four years ago, we saw it coming, so actually it didn’t surprise us from a research perspective. Now, bringing it to market, obviously there have been some very successful installations of devices. We are very excited about it. Our research agenda is working on those same kinds of technologies. Where we see the opportunities of mobile is in how people today see all these apps in the consumer world. We see a similar way of downloading apps in the enterprise setting, and a way to provide a seamless experience across the consumer apps and the enterprise apps on the same mobile device. And you should be able to leverage off the consumer world and the enterprise world. But the key thing is how you keep these two things isolated and separate, so that the things you do in the consumer world cannot in any way impact or hurt the enterprise world.

Sustainability

JF: You mentioned sustainability, and if we had another half hour we could talk about that, and changing the world. So my final question is more personal: with all your knowledge and what you’re seeing within HP Labs and all the things you’ve done to straighten that agenda and get it focused on the megatrends and all that, how is technology going to change society on a global scale in the next decade?

PB: In fact our Sustainability Research Group is looking at exactly those kinds of societal challenges. Within the Sustainable Research Group we have a project on sustainable data centers. As you know, data centers consume a lot of energy, and that’s the reason that many large data centers are located in certain regions in the U.S. where energy costs are low. But we believe in what we call the Net Zero Data Center, so that over a 3-5 year period the amount of energy it will draw from the public utility grid is zero. Essentially we look at the demand-side view and the supply-side view. In terms of supply we have electric as one source, but we’re looking at solar, we’re looking at biofuels – in fact there was an interesting article published recently on converting farm waste into energy. So there are parts of the day when you have a lot of solar energy, and you will generate more energy than your data center uses. We want to give that energy back to the electrical grid. At other times we need to suck energy from the electrical grid. So we proposed a Net Zero Data Center, and that design has been completed.

We are now thinking about how we can apply the same sustainability at the city scale. So the kinds of issues we face are resource issues about energy, about water and so-on. Some parts of the world are going to run out of clean water to drink. So how can we use sensors and actuators to do a very good plan of supply and demand to come up with a much more sustainable society.

These are societal problems. HP is working on them, and we believe that once you solve the societal problems, the business results will follow.