Home Companies Industry Politics Money Opinion LoungeMultimedia Science Education Sports TechnologyConsumerSpecialsMint on Sunday

Multi-processors will drive new computing era by 2013

Multi-processors will drive new computing era by 2013
Comment E-mail Print Share
First Published: Thu, Feb 19 2009. 01 02 AM IST

Better interaction: Craig Mundie, chief research and strategy officer at Microsoft Corp. He says computers will transform into great personal assistants rather than being mere tools that follow users’
Better interaction: Craig Mundie, chief research and strategy officer at Microsoft Corp. He says computers will transform into great personal assistants rather than being mere tools that follow users’
Updated: Thu, Feb 19 2009. 01 02 AM IST
New Delhi: As Microsoft Corp.’s chief research and strategy officer, Craig Mundie has perhaps the most powerful job in the software world. Tasked with deciding where the world’s largest software firm by revenues is headed in the long term -- defined by Mundie as three to 20 years -- he, in a sense, sets the bearing for the eco-system in the tech world. For, nine out 10 personal computers in the world run on the company’s Windows platform; any change the company makes to its products, strategy and market approach has ramifications around the industry.
In an exclusive interview, Mundie, on a three-day visit to India, talks about trends that will shape the computing world in the next decade, Microsoft’s Internet search strategy that will take a contrarian approach to the one adopted by archrival Google Inc., ad-supported business models, and how his company is dealing with the economic slowdown. Edited excerpts.
It’s not every day that one meets someone whose mandate is to look 20 years into the future. So, here’s the question: when will the recession end?
Better interaction: Craig Mundie, chief research and strategy officer at Microsoft Corp. He says computers will transform into great personal assistants rather than being mere tools that follow users’ command. Ramesh Pathania / Mint
(Chuckles) When it ends. I’ve a lot better shot at picking technology trends than picking up horseraces or how the economy will fare.
Okay... tell us about technology world will fare and the broad themes there.
One of the biggest changes will start to come by about 2013, and roughly beyond, I think you will see the influx of new styles of microprocessor (with parallel computing capabilities) machines. With that, for the first time in decades, there will be a requirement of new ways of writing software to harness the power of the hardware. It will push the threshold on the way people interact with computers. Today, when everybody talks about a computer, they think of it mostly as a laptop or a desktop and their interaction model is mostly point-and-click. But, many millions of people do not know how to use that technology and to get the benefit of technology to all those people, we have to do better than that.
I think the biggest change will be the modalities of man-machine interaction, more natural ways of the computer supporting people in many more dimensions than today.
The other thing will be that computers will become more like an anticipatory assistant. Today, it is like a tool. If you know how to use it and call it into action, it does something for you. But, if you think how great personal assistants work, their value grows over time because they know what you (as the boss) want. They can anticipate what is important for you and what isn’t. Today, computers are not there yet.
Both of these will represent a fairly revolutionary change in the way computers are used.
Tell us with examples. Will voice activated computers be one example of this?
Not quite just that. That’s still an alternative way to the click-on-file-open-it command. This will be much more conversational, much more natural user interface. If you think of it, the era of computing we’re now is defined by the graphical user interface, which Microsoft popularized over the last 20 years. Everything has been developed within that paradigm. That won’t go away; there’s still very, very useful. But, it also constrains how many people can avail of (computing) and the computer can present what it does.
In natural user interface, you take a bunch of individual technologies that we have been researching on for a long time: speech recognition, speech synthesis, machine vision, natural language processing, machine learning capabilities... each of these are individual technologies that Microsoft has been developing for some time now. Yet no one of those things has proven to be useful in the way we use computing except in an incremental way. But, if you blend them together, I’m predicting, at least, that they will create a natural model of interaction with computers that will allow us to task them in a broader range of activities and provide those services in many more people.
For example, we’ve a robot receptionist in test at lobbies at Microsoft office. It can give you directions. It (the concept) can be used in healthcare. For non-acute situations, in a world of sophisticated digital medicine, it can help like a general practioner. It can help in villages of India, Indonesia, China, Africa... I think that can be possible in the next few years.
Moving a little more short term, tell us about Microsoft’s presensce to Internet search. The reason we are focussing on search is because it is the biggest defining force, if you will, on the Internet.
Even if you think of search in the conventional sense, we certainly aspire to be one of the top providers of that search capability. We have invested heavily to do that, we have made dramatic progress in that regard. I don’t think there will be a big technological gap between us and Google in the conventional sense in the near future. But, we don’t think search as we know it will remain the way it is. Because as the corpus of knowledge grows ever larger, the ability to just take the algorithmic search that people have used in the past or gives the fastest answer is clearly not the (best) case.
We tend to think domain or task-specific searches will be increasingly more important. Also, the blending of private corpuses of data along with the general publicly available bodies of knowledge... We’re quite interested in that. We hope to have some leadership in that area.
Staying with the Internet, tell us about how ad-supported business models will change in your business. Software as a service was talked of as a model that was the next big thing but that didn’t quite take off the way it was anticipated. Advertising on the Internet seems to gained traction in a big way...
Only with consumers.
Even so...
Most businesses that have invested in software for productivity gains are not interested in taxing that productivity with having to watch ads. So, we’re not particularly optimistic that ad support is going to become a predominating way in the enterrprise software business. You talked about software as a service or SaaS. We thought of that said SaaS is software and a service.
How the business models is going to change? There are really three ways of getting paid in this online and software business: transaction fee, subscription fee and you can have a ad-supported model. Microsoft started out life in transaction. For the medium and large customers, it evolved to subscription. They just pay a fee per year and say they want all our new products and say, ‘ Bring it on.’ So, as advertising came along, it has given us an option to consider additional monetization opportunities that we didn’t have before.
Now, broadly the services model is important to us because its something that people with either transaction or subscription (contracts with Microsoft) can use it to compensate us for adding value on the things that they used to buy from us as just a licence. So we have been in the process for a few years now of really adding a service component at least as an optional part of virtually every product that the company has.
That can range from things as basic as just the maintennece of the software to things as exotic in terms of providing supplemental capabilities to the software. We are on the path to adding the service companoents to all of our Office suite products and many of our traditional windows products. So, Windows has Windows Live or Office has Office Live or Xbox has Xbox Live. Each of these things does not negate the value proposition of each device or software on its own.
You have in the past talked about two stages in the deployment of new platforms – diffusion and exploitation. Has Vista (released in XXX 2007) reached the second stage?
It’s a yes and no in that Vista is actually a continuation of the exploitation stage of the original Windows platform. Windows went through its diffusion cycle back in the 80s and then it matured and went into what I call the exploitation cycle where millions of people started writing applications for it. The platform is not reborn with each new instance (or version) of the operating system. Each of those is essentially a continuation of the next phase of that platform.
The world can only tolerate wholly new platforms about once every 15 to 20 years. For example, when I talk about this upcoming change in the maybe 2013 to 2015 time period, that is the arrival of another new platform: one that has a natural user interface, one that has a new parallel architecture in the underlying machines. That is when programmers will have to learn a new trick. There is going tio be new applications to exploit this capability. I think that will be roughly the time period we may see the emergence of a new platform. Vista has been very successful but only as an extension of the Windows platform as we know it.
How many such new extensions (or versions) before we see the new platform?
There is no guarantee that Microsoft will bring in the new platform. But we are certainly working hard to do that and if we can execute that well.
How is the new version different from Vista.
If you go back to the essence of what was the Windows platform, there is graphical user interface, the basic model of how you write applications and how people use them. That’s been a continuous thread from the first version of Windows to the present and even into Windows 7. Some of these essential characteristics are there. What gets added around it is really the means to support the steady evolution of technologies that are part of the computer at that time. There was no Internet when we did Windows, there was no Wifi, there was no mobile networking at that time.
The role of the operating system is to abstract away at the underlying (technologies) that are happening in the ecosystem around it. Version to version, a huge amount of our investment is to basically deal with that flux. Not to in some sense create a wholesale new platform. So clearly Vista had that above XP and Windows 7 will provide a lot as we move beyond the Vista generation.
A lot of work has been done (on Windows 7) particularly contemplating the arrival of some new technologies – multi-touch will be supported in addition to the tablet technologies. We will provide more automatic support for networking particularly in the home.... a lot of work has gone into the automated way of assembling home networks and facilitating the sharing and exchange of music and videos and other things between these devices without having a lot of people understand the underlying complexity.
You are an India sponsor in the top management at Microsoft. Where do you see India in Microsoft’s larger scheme of things?
We started here in 1990 just as a commercial sales activity. When I started coming here about six or seven years ago I was a proponent of thinking of India in a bigger way. Not just in a sales subsidiary and a place where we can have development done in Hyderabad but to basically think of India as a rapidly growing economy and one that could house more components of Microsoft’s business than would have otherwise been the case. And, today India has the second largest population of employees of any country in the world. Onluy the US has more people working for Microsoft.
Its become a much more diverse capability. We still see it as a expanding commercial opportunity. Still the bulk of our revenue today comes from our traditional product lines. We don’t have that much revenue yet in our entertainment and devices category. The online services business is still quite fledgling here.
We are concerned about India as we are about any other country in this general economic malaise but I personally think that if India and China, if at a local level are well managed by the government, they should be able to recover in this general economic environment (faster) than western Europe and North America.
Does the slowing of the environment for Microsoft and the laying off generally affect your research turf in any way?
There is really no part of the company that we won’t make adjustment to. And the reason is Microsoft’s top management view of the current global economic situation is that we are moving to a permanent reset of the economic level on a global basis. We don’t view this as a bounce where its just going to descend and then come back quickly to where it was.
The reason we did this first round of layoffs and said we would do more restructuring is that we want to be plan-ful about it. But we do intend to alter fundamentally the company’s cost basis to reflect what will be a stable bottom in terms where we think the economy will stop globally and then start a slow reconstruction after that.
What would constitute those step down measures?
We basically laid off about 1,400 people worldwide last month and at that time we also announced that over the next 18 months we expected that we would remove as many as a total of 5,000 – so another 3,600. Each business group is now making their own plans, discussing their own situation and deciding how they will bring the company gracefully down this new structural level and that will be an ongoing process.
Comment E-mail Print Share
First Published: Thu, Feb 19 2009. 01 02 AM IST