Email us


This area does not yet contain any content.

Blog Index
The journal that this archive was targeting has been deleted. Please update your configuration.

Directing Clouds

It isn’t easy to interview Fausto Bernardini, the director of Cloud Services at IBM. If he is not flying to Australia or China, if he is not on the phone with clients or with his team, he may be riding his bicycle or skiing down a mountain depending on the season. Born in Rome, a degree in Electrical Engineering, a doctoral degree in Computer Science at Purdue University, he started working for IBM in 1997 as a researcher and never left.

Help us understand the importance of Cloud Computing.

Cloud Computing is a new way of delivering IT (Information technology) services to consumers and enterprises, and it is about having centralized large scale server farms, and large amounts of storage that are rented out as a service to be used by many people.

Many of the underlying technologies are not that new. The concept of having a virtual server, that is the ability to carve out a piece of a physical computer to be used as a server, and partition that machine in many virtual ones that could be used by several companies at the same time, goes back to mainframe days.

What has happened in the last five years is that first of all commodity hardware has become powerful enough to run enterprise applications. Then technology was created to virtualize it. And finally, a number of companies in the “Web economy” were also born on the Cloud. They started building their IT (Information Technology) from scratch with this new Cloud architecture in mind and they demonstrated to the rest of the businesses that you can run planetary scale, multi-billion-dollar businesses using Cloud.

That, of course, got the attention of everybody in the industry, and people started to realize that Cloud is a terrific new model that provides a number of benefits. What it does is very valuable because it moves the IT capital cost. Normally companies would build their own data center, invest a lot of money to do so, and now they can shift that to an expense. You can rent as much capacity computer power and storage you need at any given time, and you don’t have to pre-commit to a certain quantity. So, if your business is doing well, you can simply rent more, and if your business is in a period in which growth is not meeting expectations, you can also scale down your cost.

It’s a much more dynamic way of doing business. It’s also a much faster way to obtain the resources needed which allows for more experimentation and business innovation. Small companies can now have a global reach because they can get on with a global Cloud provider, and they can replicate their application in many data centers around the world without investing a dollar of capital. It is an incredibly powerful new tool.

What stage are we in?

Clearly we are still learning. There are significant proof points or large companies including many internet household names that run in the Cloud, some working with IBM Smart Cloud, and new companies are added everyday. However, if you compare that to the total IT spend, it is still a very small percentage. 

I actually predict that a significant percentage of IT spend will go to Cloud. I don’t believe that everything will become Cloud because we are still running applications that are sixty years old on mainframes. And we will probably continue to do so for decades to come. So, we will have a hybrid world but Cloud will take a very prevalent part of it.  

The research and technology behind it is still evolving; there is a lot of investment being made, there are many interesting start-ups that are creating innovation, there are open-source activities, there are standard bodies that are creating open API’s (application programming interface) and standards so that different Clouds can be interoperable. There is tremendous innovation activity and investment in this area, and we have only seen the very beginning of this transformation. Definitely it is not 100% mature; however it is serious enough that big businesses can run on it today. I think that in the future Cloud will become the default choice for a company.  

You will have to have a very special motivation to not be in the Cloud; the exceptions will not be in the Cloud and the standard will be Cloud. The exceptions may be companies or institutions that are too old to move, with very critical or particular workloads or applications that require a very specialized type of environment. Perhaps the ones with the most sensitive documents or financial data, but in general the majority of applications and data will migrate to Cloud. 

I think that the economic underpinning is very strong. With Cloud what happens is that I, as a Cloud Service Provider,  can invest a lot in innovation and in building up capacity, and I can spread that investment over a thousand clients. Each one of those clients cannot match the level of investment or capital needed. The economic benefits are all in favor of Cloud, and I think it is unavoidable that it will become the de-facto standard.

Yet, people are scared of Cloud because of security issues....

In part they are right, meaning Cloud is too new to have demonstrated that it can be trusted as much as other models of computing. But it is simply a matter of time. People say: “do you have a track record to show me?” Because it’s so new, we don’t have a lot of track record. But for most people it is not lack of security they are worrying about, it is more lack of control.

When a CIO (Chief Information Officer), and I talk to CIOs of Fortune 100 companies all the time, asks me: “what about security in the Cloud?” I answer: “what about security in your data center? Are you sure it is that good?” The CIO usually laughs and gets the point.

Data center security is a difficult proposition; it takes a lot of investment and practice and process discipline. If anything, it is easier to do it in the Cloud because one single large provider with standardized processes can secure the same data center architecture all over the world, much better than a single CIO can secure their own data center which is made of a collection of heterogeneous systems. It is just a matter of time, in part Cloud Security needs to mature; in part it is just getting comfortable with it; in part it’s people starting to get more at ease with that loss of control.

How about the global legal issues of data protection and privacy?

There are different aspects in that. There is an aspect of policy and protection of privacy and private data that is very important. The US, the EU and other countries around the world have been working and revising different privacy bills that cover this new sharing of information. IBM participates, just like many other large IT vendors, to these types of forums and activities, standard bodies, and policy bodies. But it is in its very early stages.

If you shift from Government to a large private company that has IT security policies that were written before Cloud, and that are often inappropriate to handle Cloud as a model, you know that they need to be updated, something that with security does not change quickly. Essentially you know that what you have now works, or at least you know what doesn’t work, and whatever you change is going to introduce more risk. People are reluctant to change those policies. 

The EU (European Union) has very stringent rules; they don’t allow private data to move from country to country unless the data goes through a number of controls. Some of that is appropriate and definitely good to have to protect consumers and citizens, but, again, sometimes it is an overshoot that hinders new development and innovation. The EU is taking it very seriously, and we have an open discussion with policy makers on that. The same thing happens in the US with the Patriot Act, something that other countries fear because it gives the US the right or the perceived right to go and seek data in other countries.

These things have different interpretation, we are moving the first steps, and things are still a little bit confused. IBM is very active in this debate with policy makers globally to make sure that rights and privacy of citizens and customers are protected but also that technology and innovation can take place.

What happens when you present the Cloud in countries that haven’t adopted it or don’t have the infrastructure to do so?

I am an advocate, an evangelist of the Cloud. I am trying to push the envelope of what people want to do today when this thing is still not completely understood or trusted. That’s why I like my job, we are breaking new ground.

Cloud is the future but you said that the Mainframe is still a significant part of IBM.

It’s kind of funny. Mainframe has been called a dinosaur for many years, but it is a dinosaur that won’t become extinct. IBM still has a very active and profitable business around mainframes, which by the way have evolved. In many cases a mainframe is really a private Cloud. That’s the way companies use it. It is a machine that is capable of running tens of thousands virtual machines running Linux or standard operating systems. It’s a physical machine, it sits in a client data center, but for a large company it is used across different business lines as its own Cloud.

There is a concept of a public Cloud and a private Cloud. A public Cloud is a third party service provider that is investing all the capital to build the data centers so it can rent them out to many users.

For some large companies it also makes sense to have a private Cloud; they are their own service provider, like a bank or a large manufacturing company. They still get the benefits of Cloud in the sense that it is a dynamic environment in which service and storage can be allocated on demand. But it doesn’t have the economic flexibility benefits of a public Cloud, when you are relying on somebody else putting in the capital. With a private Cloud you are paying for the hardware.

If you are large enough, a private Cloud may be a better model. If I am going to visit a city for a few days, it’s better for me to stay in a hotel. If I know I will be there for 3 months, maybe it is better for me to rent an apartment, and if I know that I will be there for five years, I might be better off buying an apartment. Imagine the Postal Office: they have a very stable predictable business, they know how much storage space they need. It may be preferable for them to buy it, put it inside their data center, and have full control of security. That is essentially what the mainframe has become; it is a private Cloud and it has all the latest technology. By no means would I define a mainframe as a technology dinosaur.

It is still a great business for IBM but it’s only a part of our business. IBM makes about half of roughly one hundred billion revenue in services, and the rest is split between software and hardware. It sees a lot of growth in software and has a very successful enterprise hardware business.

You worked in research for many years: is it still an important part of IBM?

We continue to be at the forefront of corporate research. IBM has established research centers here in New York (Westchester County), in Switzerland, India, Australia, Japan, Ireland, Israel, China, Brazil, Texas and, of course, the West Coast. 4,000 scientists work in these labs essentially inventing tomorrow’s technology. We are also investing greatly in growth-markets, we just opened a research center in Kenya. We are expanding globally both in established and growing economies because research was and is a pillar of IBM’s strategy.

While you were in research, you built digital models of Michelangelo's Pietà in Florence, of the Throne of Tutankhamen and of the statue of Pharaoh Khafra the II in Cairo. Was it brutal to go from that environment to the cold business side of IT?

That was a cultural shock. It took me probably six months to get comfortable in the business world. It is more fast paced than research, in research you ponder a point, you don’t make a decision until you have all the data available to prove something. In business very often you have to go with your guts, you have very incomplete information but you still have to make a quick decision. That’s a new skill that I had to learn.

The thing that I loved the most about research was the intellectual exchange with my colleagues. You find yourself in a building full of bright minds, all very passionate about whatever research they are doing at the time, and a lunch conversation is like a documentary from a science television channel. That’s probably what I miss the most. I felt very lucky to be part of the research team. 

But I like change. I was in research for 17 years, and I needed a new challenge. I was very curious about the world of business. I spent the last 3 years in research developing Cloud technology, and I wanted to see what that meant in the real world. I wanted to follow Cloud from the labs to reality, to the market. That was incredibly satisfying.

My time in research feels like a past life, but I try to keep current. I am still contacted by researchers asking for a paper or an image from my previous work, and my articles are used to teach classes in computer graphics.  I was always interested in the application of computer modeling, computer shape acquisition to art history. Studying and really understanding our past by visualizing it. I still find it fascinating.

Is it possible for a country like Italy to rely on developing IT start-ups as an economic strategy to fight the recession?

Italy has the creativity and the learning institutions that could support becoming a leader in the field. What I realized when I was 30 years old - after I moved to the US - was that clearly we lack a culture of entrepreneurship and the underlying infrastructure to make it possible.

Silicon Valley is not just bright minds out of Stanford or some other great university coming up with an idea, it is also the world of venture capitalists that are ready to bet millions in new ideas. We all know about those ones that succeed but for one that succeeds there are a hundred that fail. That’s all part of the game.

That culture of risk and experimentation allows young people to do something exciting, to work hard and start new businesses instead of looking for a steady job they may not even like. It’s an old Italian mentality and it will take time to change it. But in order to change it, the Italian Government should simplify bureaucracy so it won’t be so difficult to create start-ups, and provide incentives. 

Then there is a role for the financial industry to believe in this change and provide funding. Nothing happens unless you have the inventors, innovators, and investors. It requires the right environment to flourish but also a change in the mentality. Fortunately, the world of start-ups is far from the world of corruption so that it is not an issue. It’s all about good old competition. The best idea and the best execution win.

Many Italian-born people who live in the States do not consider themselves Italian-American and can’t wait to move back to Italy. You have a wonderful son, Tommy, who was born here and is first generation Italian-American. How about you? 

At the moment, I am not thinking of moving back to Italy, it may be an option when I’ll retire. I live an American life, I am an American citizen, I work for an American firm, and I am definitely an Italian-American. I have roots in Italy, my family, my friends, my wife’s family, we go there very often and I enjoy every moment. I am very much connected to Italy, but if I ever moved back it would take me a lot of time to readjust.

***The opinions expressed here are Fausto's personal and do not necessarily reflect those of his employer, IBM.


December 8, 2012