Interview with Anil Valluri

Anil Valluri recently took over as Sun Microsystem’s vice-president and Managing Director for India operations. He joined Sun Microsystems India in 1999 as its director – Systems Engineering and CTO, responsible for all technology, product and solution consulting. Valluri led a team of IT architects, consultants, product specialists and solution managers in Data Center, Data Management, Java Web Services, Identity and Network Computing. Before being appointed vice-president, Valluri was heading Sun India's Services Business, which spans Support Services, Professional Services, Learning Services and Managed Services.

In his first interview after taking over the new responsibilities, he shared his priorities and discussed various Sun technologies and its future in a talk with Srinivas R of CyberMedia News.

Recently you took over the Sun India responsibilities. What are your priorities for Sun India and what changes you would like to bring within Sun India?

We don’t have to bring any changes. The idea is to have continuity of lot of good stuff that we did all these years. We are only going to fine-tune and tweak some of those things, which we always wanted to fine tune. And also, there is the market, which is always changing and dynamic. You have to continuously adapt to the changing market conditions. So the changes that we will be making would be to serve the customer better, to have lot more partnerships aligned to again add the customer requirements and also look at how we are developing the employees. So clearly, my focus would be on three constituencies, which are Customers, Partners and Employees. So those are the three areas that we would be looking at in fine-tuning, and it could not be wholesome changes but these will be basically tweaks and fine-tunes.

You are referring to the market changes. What do you mean by this market changes? And what are you referring to?

The markets have changed significantly over the last few years, in the sense that there is a huge explosion of the SME market. That is, small and medium businesses are the ones who are buying today. Large enterprises have bought and have got two cycles of purchases and they have been matured enough to handle lot of their businesses. So, those enterprises have invested and they may not being growing too much; whereas, the next year of enterprises or customers are the ones who are going in for IT, adopting IT or going to the version 2 of IT. So there is a huge explosion of market around for small and medium businesses. There is a lot of focus on government. Earlier government has been a facilitator to lot of IT purchases and requirements. But today, governments themselves are big buyers of IT, because they are investing in various social improvement projects, they are on lot of citizen services projects and so on. Whether it is income tax or customs or passport, or whether it is infrastructure, whether it is a land record forms and so on and so forth. If you look at any economy, which is developing or developed, government is a large buyer except in India, where the government has not been a big buyer. That has to happen in one of these days.

So we have to really be aligning with government, working with government and Sun brings lots of good stuff to government in terms of Open Source, in terms of Java, in terms of Open Office, in terms of some of the standards that would open solaris and so on. And there is a large community of developers that government again can bank upon, which is again on Sun’s technologies to go and do these projects. Sun plays a very large part at the technology level, at standards level and as a vendor as well. So government is a big buyer that is another change that has happened. Telecom is an exploding market. And everybody knows about it. And India continues to be an exploding market for Telecom. Subscribers are only getting added by the day and by the week. So, we have to focus on that. Sun has been a strong player in Telecom. We will continue to focus on it. Financial Service market again is a big exploding market. Hundreds and thousands of online customers are getting added. Banking has moved away from counter-based banking to ATM-based banking, to electronic banking. There are changing patterns, so there are changing requirements and dynamics. We are going to continue to address that space. Retail, is a new market again. Worldwide retail is not a big market because it is a saturated market. In India, everything is new.

We are building all the stuff. Retail is going to be a reality because consumers demand services of retail markets. So if it is not X, somebody Y will have to serve that requirement. We are focusing heavily on the retail. We have had very good successes with costumer’s initial retail deployments in India have all been on Sun. So we will continue to have that. Those are the markets we will focus on, those are the customer patterns that we will look at. Sun came out with whole series of products around x86 product lines. We offered windows, we offered Linux, and we offered Solaris. That opens a new market to Sun again. India is a very big x86 market globally for large players. So, for Sun with strong brand it has, it is not going to be tough on us to take few percentage points of market share away from somebody else. And that would present again a huge boost to our revenues. So we will do all that is required to have x86 being offered to the market place. These are the broad ones that we are looking at, or these are obvious ones but these are sometimes simplest ones are obvious ones.

During the dotcom era, your tag line was “we are the dot in the dotcom”. Now it has changed to the “Network is a computer”. So, in this era of social networking what role does Sun play?

A very big role, I could say. Network is a computer is the mother of all our taglines. It is always been the overarching tagline for Sun from 1982. So we use network is a computer as a main mission and vision to address the market. Whatever we do is to enable network computing; whether it is technologies, innovations, people products all that stuff. What we do at different points in time, use the relevant phrase to associate that network dotcom to the market place. For example, “We are the dot in the dotcom” is relevant in the dotcom boom time. We have had another tagline called “We make the net work”. Currently the tagline is “It’s the age of participation”. We will use it to relate to people, to get them to understand what we mean, these sub taglines to associate to the big tagline. So, currently the one that we use is the “Participation age and developing communities”. The fantastic question what you asked about this social networking is all about participation. Earlier, the web was something where one could seek information, read the information and find out or do shopping or whatever.

Today’s network is all about participating; you write back, you upload stuff, you upload your videos, you chat, you have blogs, you have all other stuff. Social networking is just one part. There is government networking, there is business networking, there is commercial networking, there is statuary networking, there is crime networking, etc. And that is all social in a two-way networking; this is what we call the Web 2.0. That is a huge boost to the industry at large because these all help by protocol slab, the lamp structure we call that. And we have trying to acquire properties around LAMP, which is if you take Linux, Apache, MySQL and the PHP. We have gone and acquired MySQL part. Why is that, because we want to be relevant in the lab space. The L base which is the Linux base, you are also offering an alternative which is S base, which is the open Solaris, similarly at Apache we have glassfish, at P level we have some thing else so we are acquiring properties and developing technologies to replace some of those once, which are probably what is used as default. With better value and better bug-fix capabilities and assured support and so on.

While it is open source we give assured support. So Web 2.0 is very important for Sun and very relevant because we offer a lot of infrastructure. Lot of infrastructure is used in Web properties are all Sun-based. And that also presents another interesting case, which is lot of business case with huge exploration in storage, huge exploration in infrastructure, huge exploration in networking protocols and software development, application servers the whole stack, data bases, so there is commercial side to it because there is a exploding requirement; because if you and me are uploading videos, obviously they have to store it somewhere. If I have to use free web mail somebody is offering, somebody is putting all those servers.

For Sun, Web 2.0 is extremely important; networking is extremely important and our tagline is always, “We want to build communities to enhance the age of participation.” This is the truth, which we cannot ignore. It will continue to grow on its own and if you are part to it, you will benefit from it. If you are not, you will be left out. Sun being a network provider for ages, of all the things we do -- whether it is Java, whether it is Glassfish, whether it is MySQL, whether it is Open Solaris, whether it is open office, they are all integral part of that networking or social networking or Web 2.0 networking style. For Sun, it is very relevant that we continue to develop that market. We encourage developers and we have 700,000 developers in India and they are all the ones who are creating the networking sites.

By the way, your CEO (Jonathan Schwartz) is very popular in the blog world.

Absolutely. I think that is again symbolic of what we represent. We are not a company, which is just looking at standard business, enterprises businesses and so on and so forth. We are very, very strong on the enterprises. We become very strong on the enterprises side because we are very strong on the networking side. But lot of that game played to our strategy, which is why we bet on Internet big time; we bet on bandwidth big time; we bet on user interfaces, which are web protocols. We did not bet on desktop is the strongest piece. We always bet on the best half should be the lightest piece of the architecture. Way back, I remember, in 1999-2000, when we said network is going to be ubiquitous everybody used to laugh, in India particularly. When we used 9.6 kilobyte phone link.

Now, if you don’t have network access at a good speed, then something is wrong somewhere. So ubiquity of network, bandwidth, unlimited bandwidth -- those are the ones, which we are taken those bets on and they all turned out to be true. We always took those steps, bets in advance and we invested in technologies and created those protocols. If you look at the two biggest communities for developers, we own half of that community the other half, probably belongs to Microsoft. There is no other competitor who can claim to have developers on their side. Which is what make Sun so relevant in the market place and so viable in the market place, because of zillions of people are using your technologies, and are making a career out of that technologies or making applications out of your technologies. Then there is something good about you and something right about what you are doing and others can’t claim to have all that. They could be great service company; they could be great printer company; they could be great whatever, but not on the developer side.

Globally how many Java developers do you have?

Globally, we have about 3.5 million developers on Java and we have over 5 billion devises, which uses Java. These are all very big figures in India. We have about 700,000 developers around Java and Solaris. There is a lot of buzz around green computing and how is Sun planning to balance environmental issues with scaling computing like power consummation and so on? Sun is very much in the forefront of green computing, I think sometime back we did a campaign around eco-computing, green computing, using our black box and all that stuff. But the most important piece is we are not promoting green computing because it is fashionable and it is the right thing to do. We are doing it because our technologies are truly green. With the stuff we did around chip-based multi-threading which uses the CPU cycles when they are ideal to do other things. We have reduced power consummation phenomenally; we have increased the performance significantly without increasing the power requirements. And we have also compressed the footprints because they are just very small chips with very low power requirement, which you can package it to 1u and 2u spaces. We have a metric called SWAP (space, watt, performance). Using SWAP rating, we differentiate ourselves in the market place by atleast between 5 times to 10 times using SWAP as a measure.

The point is, it is not something, which you are doing it because it is important and it is something what market is looking for. But our products are actually very very green. To give you an example, our Niagara-2 processor and the system runs at 70 watt of power and it gives you 32 threads of performance. Almost having 8 cores. Hugely high performance, very low power requirement, and very low head deification thereby, which means also very low cooling requirements. So, green is that plus we also are looking at designs of data centers, which are very innovative. In terms of cooling, only those specific hot spots and not cool the entire data centre for some x degrees of Fahrenheit. There are very innovative cooling methodologies; we are using very high efficiency power supplies in all our systems and something like 98 per cent deficiency in power supply designs. We use very innovative cooling methodologies for all our fans and air turbulence and so on. All of this is bits and pieces. But the big piece is actually when you stat them and drag them in large data centers, the magnitude of power consumption is enormous. For example, to test showcase this we have actually setup the data centers in Santa Clara and in Bangalore, where we demonstrate how much of efficiency and energy we are saving, etc.

In fact, our Santa Clara data centre has won a lot of awards in the State of California, which I think is probably the most environment conscious State in the world. Various tight norms on environment and they have awarded our data centre for being the most efficient. In fact there are figures available as to how much of carbon footprint and how much of all other stuff is saved. For Sun, green is very important because we have real technologies which facilitates that, we also have very innovative architectures which will add to that in-terms of cooling designs, in-terms of power designs, in-terms of data centre layout and so on. We want Sun to be known not just as a good company as a great company. One of the greatness of being a great company is we should be green, we should be ethical, we should be meeting the requirements of the society, all of this. Our Bangalore data centre, where we have used all of those technologies, has been awarded the best data centre by PC Quest, in terms of the energy designs.

What would be your offerings apart from MySQL servers and Java?

Sun is a company based on innovation. Developing IP is our differentiator. We have to differentiate ourselves in the market using an IP. Our differentiation is not going to sell somebody else’s IP. Therefore, we will continue do our offerings are obviously Java is a good example that you have given. We are very strong on Solaris and these are flagship technology. We have offered the new set of technologies, which is the MySQL in which we have acquired and offering it to the market place, very very popular web properties for data base requirements. Then we have another big differentiator -- is what I mentioned chip based multi threading designs -- where we have innovatively used micro processor design to get deficiencies to an order of magnitude higher than what it normally is by effectively making use of the CPU during idle time. When CPU is doing processing it waits most of the time for memory to come back because of memory latencies. So while that is happening you want to multiplex it and use it for something else.

So while this wait is going on, we do something else, while another wait is going on, we do something else. Having said that we create multiples of threads capability around the processor, the good example is the Niagara processor, which has 8 cores and 4 threads per column and does 832 threads of computing on one single chip, which is of 70 watts of power requirement. We are improving on the designs; we are increasing it to 8 treads per column, we are increasing the number of course to 16 course per chip and all of that. Innovation will continue to happen. This is under system design. On the higher end, we have also come out with SPARK version 7 chip, which is again a quad-core chip, a very high performance chip for the large scale computing requirements for banks, telecom companies and so on.

What is the future of open office? There are lots of downloads happening but what is the revenue model?

Open office has been an effort from Sun to the community to give them an alternative to the standard offering that they were used to. So this has initially meet with lot of resistance people didn’t want to try some thing new they got used to some thing. But I think, now more and more customers come and tell us, that their defector standard in their office is open office. And they are all doing business commercial enterprises are running open office in their organization, I know of so many banks use open office as their office productivity tool. So what gives us we keep telling, Jonathan keeps saying all the time, our biggest challenge is been to stare on your face on the desktop, people know or don’t know Sun because we been in the datacenter for the most of the time. By doing things like Java on your phone, by doing open office on your desktop, Sun is in your face because you have been using it and you see Sun logo most of the time. When you use some of that, you will see Sun copyrights and all other stuff. This is what we are looking at, adopting technologies by people which Sun wants to. Why do we open source? We want people to use our technologies. We are expecting that we will be able to make revenues out of it by selling them service offerings. We are a company, which develops technologies and give it to open source.

Therefore, costumers can be very sure of the quality of the offering. It is not developed by a bunch of people who came together for that purpose and disintegrated after that. But it’s offered by a company, which has put engineering effort and sources behind that and continue to update it and continue to offer bug-fixes, patch-fixes, etc. Whenever you want to really take service, you can go to them and take that service. The same is the case for a bank, or a large telecom company. While they use open source, there is no compromise on the quality of the offering, because it did not come by in a day’s effort. It’s a continuous effort and whenever they want to take support, they can go ahead and take that support from Sun. The last thing that we want large enterprises to do is to go to a chat room when their bank is down and find out how I can fix my bank or my data centre. You don’t go to a chat room; you go to a vendor who knows how to give you a short support. That is the thinking behind Sun’s offering. We will get our technologies adopted by giving it away into the open source and monetize it by service offerings or increased infrastructure requirements by increase in usage of all of these web, Internet, networking and all other stuff. We eventually sell more servers, more storage other stuff and create the demand for ourselves.

A good example, which we always give that how do we monetize for is Z file system that we have. Z- file system is an open source. So many people have downloaded, used it and become so popular. And then we offer that because of the nature of the goodness that the Z file system offers in a product called Thumper. It is a X4500 product from Sun, which is a combination of a file server and a system. It’s a storage and server together. It’s a unique hybrid between a storage and server. It’s very popular because it has server, it has disc and it differentiate itself because it has Z file system on it. It does all the functionalities of a high-end storage box. It is very good for using in data marts and data warehouses and so on. Z file system first we made it popular and made it into an offering packaging to get onto the server. This is a good example of how we monetize. We are very steadfast in our belief on this one. More people use our technologies in one form or the other. More is our relevance to the market. More relevance then there is a good chance and good opportunity for us to make money and go and do more business. If you are not relevant; then there’s no way that you can go and make money or business.

How many open office users are there in India?

I wouldn’t know. Actually we don’t track because there are millions of users. We don’t track that anymore. That is the level of true open sourcing that we have done. What we are looking at is how do we get them to buy service contracts from us. And we are also not targeting the non-mission critical users. Mission critical users are the ones who will be going and approaching them. And it just starts open office, the open Solaris downloads. It’s big for us. Over 11 million downloads and more than 70 per cent are non-Sun downloads. So we have a good opportunity to go and sell service contracts on open Solaris with HP and IBM systems running. So, our attempt is to go to those people who have downloaded open Solaris and give them a good value proposition. Those who need it will buy it. And that’s good enough. We are not expecting large numbers may be students; large numbers may be education institute may not need it. Some people will and we know because downloads are done from Sun sites, we can go and talk to them.

What is Sun India doing on the mobile space? There are more than 5 billion mobile devices that use Java. Apart from that what are the other offerings that you have?

Again the attempt is to create demand on one side and monetize on the other side. By having so many users use Java and get into those developers space into your hand phone and into the devices, which are been used by individuals. On the other hand, there is a demand that is fueled by this usage of increase servers, increase storage, increase billing systems, increase database systems all that we will continue to go and sell. But on this side, how do we monetize? We may be monetizing indirectly again through adoption of those people who are developing those devices coming and licensing with us, because they license Java footprint. But the other way, it’s also enhancing it from a technology basis. We have come out with something called Java FX, which is rendering scripting tool for Java, which is agnostic to what is the device it is been used on. So it does not care whether to use it on a mobile or on a PC.

It renders to the screen that is being used. So, that scripting language we use is called Java FX. It’s a Java language we have come out with, so that it makes life easy for the developers to develop applications around mobiles. All of the gaming, for example, is a huge source of revenue for Sun because mobile gaming has huge requirement on the backend. So, do we monetize that? Yes, we have a good shot at monetizing that piece. We may be competing with others as well, but we have good shot because we are fueling the man. On a standalone basis are we going and selling to consumers? No. But are we going and selling to handset manufactures? Yes, of course. There is a Java license that is associated with that. And then we will also monetize on the other side the server side, where we actually fueling the demand.

What are the other offerings like Java FX. Is there anything else?

Java FX for the mobile part is the single biggest offering. But other than that, we continue to give tweaks for the fixes and the next versions and so on. But that’s about it. I think Java FX is significant.

What is happening at Sun India R&D?

The projects are not revealed obviously -- engineering projects, nobody would tell you. We keep working on different projects. We have talked about optic fibers as back planes for communications within the system. We have used the proximity computing that we are talking of user’s capacitance as a means of communicating between two chips placed next to each other without any wire; just the air is the medium, lot of innovative stuff. But those projects, which we have shown to or talked about it to the marketplace. We are also doing a lot of work around chip-based multi-threading, the next generation of chip-based multi-threading systems. Coming back to Bangalore R&D, Bangalore Engineering Centre is a site, which is a part of our global engineering organization.

They do work on basis of collaborative engineering means some pieces will be made there, some pieces in California, some pieces in China and so on, based on where are engineering centers are. Some are out of Russia. They all put together managed by project managers sitting in different places. Project leaders could be somewhere else. So, we will not be calling, this is a project or a product developed out of Bangalore or this is out of Russia. Sun does not do that. Sun has highly distributed sales force, the technical force and the engineering force. Having said that, it is unfair for any engineering centre to say, this is my project. Lots of others from other places have contributed to it. So, India as a centre has been working on Solaris, on Java, on the application servers, all of these on the Java enterprise stack.

Sun India has grown 30 per cent last year. What are your projects for the next year?

We would like to do more than that. The point is, we would always look at the market, what it is in a year and are we going faster than the market or not. If we are growing at the market rate then we are not doing justice. If we are growing slower than the market then we are losing market share. The idea is always to gain market share points by growing faster than the market. That is the idea. Current IDC figures talk of 20 per cent as market growth. So, we want to grow more than that. That is something, which we have been doing for nine years now. We have been growing faster than the market. The idea would be to continue to beat the market growth. As much as possible, we don’t limit ourselves in terms of -- okay, we should grow only so much and no grow this much. We want to grow more. And we have the right offerings to the market place, which I clearly mentioned to you, which are around services, around x86 offerings, around the high-end enterprise servers, strengthening our presence in the telecom space, the banking space. Retail is a new area, commodity is our new area and we have covered lot more SMB spaces with big focus on Java. So, all of those pieces are the ones we are doing and investing, just to be ahead of the market and we will continue to do that.

How is your emerging market region and going to benefit the customer?

The emerging market’s region is a great idea for Sun. Because what we have looked around and said there are these traditional markets which are the US, and the Europe and the Japan, Australia and so on. These are developed markets. And there are markets which are new, typically big countries and some of those other ones, which are Middle East and so on are big growing markets. These markets have unique characteristics. If you were to lump these countries into administered areas, which are administering or managing the traditional developed markets, it would be very difficult for them to associate with the requirements of the emerging markets. So, we have carved out these set of markets. We called it the emerging markets region. Because they have a unique set of characteristics of behavior, of buying, of decision making and so on. And we have lumped it together as the emerging market region. What are those characteristics? The characteristics are they are very price conscious. They are very open source oriented. The government does not want to buy and get locked into any vendor’s technology. They want to be independent. Most of these countries China, India and all of those, they are fearlessly independent in terms of what technology government needs to adopt, so that they are not locked into any vendor’s products. Tomorrow if something happens, they don’t want to get caught into that. So, open sources is a big thing. Geographic coverage, all of these emerging markets have vast geographic coverage.

The coverage sould be different from emerging markets v/s developed markets. We wouldn’t be able to cover, with just of a few set of people. This productivity norm’s that we apply outside is not good for the emerging market. Having said that, we need investments in these emerging markets, you need a lot of air cover for working with governments. We need lot of air cover to develop the developers. You need to go and popularize your open source offerings. So, investment again is a key characteristic of these emerging markets having such divergent requirements from compared to traditional markets. It is not possible to do within the same regions; we have to carve out separately. That is what we did. We look at these things and make those offerings relevant to the market place. That is the emerging market region: we have entire Latin America as part of that. We have EMMA (Eastern Mediterranean, Middle East and Africa), we have the CIS states, we have India and China as part of emerging markets. So huge entry, if you look at the map, it almost covers 90 per cent -- big area and time zone spread across around some 12-hour time zone spread. It is interesting and this is the investment that Sun is making and this is the priority for Sun.

Is Web build-out for customers an important strategy for Sun?

Of course, there are a couple of trends that we see. Web build-out is a very very important trend because everything and anything is offered over the web today. If you are doing CRM, if you are doing HR, if you are doing ERP, everything is web-based. It is not that you can do anything more in a traditional way. That is why build-out is happening. And social networking piece is another great part of that web build-out. Whether it is Facebook, whether it is Myspace, whether it is LinkedIn, all of that is a huge phenomenon there. So web build-out is a market that we look at. There is another one that is happening. HPC (High Performance Computing) for cloud computing. For offering it as a utility value.

So traditional HPC has been technical-oriented HPC and now it is commercial HPC for doing a lot of simulations on your financial models, the stock markets and all kinds of modeling is done using HPC for commercial applications, commercial purposes. So HPC is another market for us. The other market that we are looking at is the enterprises, where enterprise applications would continue to be there and we would continue to grow. And this is traditionally a strong market for us. This is what we call the commercial applications or the enterprise applications. So, that is again we will continue to focus. Those are the three markets that we are looking at from market visibility.

Scot McNealy has said that the cost of exit for the costumer is cheapest in Sun’s technology. But isn’t the cost of entry equally if not more important?

Traditionally, people look at the cost of entry as a measure of purchase decisions: which is the cheapest that I can buy? What Scot McNealy brought to the table is a completely different thinking, which is what is the cost to exit. We call that low barrier to exit. Sun offers low barrier to entry and low barrier to exit. If you buy a product, which is very low prize or free and you cannot come out of it when you want it. Then that would be called high barrier to exit. If you can exit very easily and move on to another technology or port it onto another technology, use it seamlessly on another breath of systems or technologies then you would call it low cost of exit, low barrier to exit.

Sun offers that. Sun not only offers low barrier to entry, because from a prizing stand point and all that we would be as competitive but the unstated which customers ignore most of the time, is low barrier to exit. Because ours are so open and so standard in terms of the protocols and then these specifications that we build on, they are so standard that you could just remove Sun and put somebody else as well. It wouldn’t make any difference from portability prospective, from application prospective, so that is what we mean by low barrier to exit. And with Sun, this is a unique differentiator.