He likes to call himself the oldest. Rick Rashid is the only Microsoft executive who has been in the same job for the last three decades.
Rashid, senior vice-president at Microsoft Research, who is in India as part of TechVista, Microsoft Research India’s annual research symposium, talks to Mint about the challenges in research, technologies developed in the company’s five research centres and the future of computer science. Edited excerpts:
Research and development is the most integral part of any technology company. What are some of the challenges a company such as Microsoft faces in the age of changing technology?
The most important thing you can do in research is hire all the right people and give them the right environment to work in. It can be particularly difficult at times and there is always a danger, especially when you get researchers trying to chase technology or business trends. You never quite catch it. Business climate changes rapidly and research takes years. However, if you are doing great basic research, you will be sitting out on the road when the car gets there and that’s where we want to be. We have often made investments in areas long before any part of Microsoft was even concerned about it. We invested in computer vision long before Microsoft even cared about it. Similarly with graphics and media technologies.
Rick Rashid, Senior Vice President, Microsoft Research
The reason you do basic research is to be able to change rapidly, so when there is a new competitor, a new technology or a change in business plan, the reservoir of intellectual property, technology people and intellectual infrastructure you have build will help your company adapt more rapidly.
Can you give us examples of this?
In 1992, right after Microsoft Research started, we began working on a set of optimization technologies for software when it wasn’t something the product groups needed at that time. By 1995, we were close to shipping Windows 95 and Office 95, and the problem that we had then was that the amount of RAM (random-access memory) in personal computers in that era had not increased very quickly, partly because of trade issues in Asia, so RAM prices were unusually high. Our competitors in the office product space did not have that technology, so that gave us a huge competitive advantage during our critical transition in the industry.
Our entry into the search business is again another example. When prior groups made the decision that they were going to be directly competitive with Google and Yahoo in that space, we were able to work with them and help them get an entry into that space in a much shorter period of time than any of our competitors were expecting. In some cases, however, it is more how important is it for you to survive.
You have a budget of about $250 million (Rs982.5 crore) a year on research. If we were to look at the amount spent on Microsoft Research in its entirety versus what’s come out of that, does the value match up?
Bill (Gates) and Steve (Ballmer) have both said in public that they believe that our investments in basic research are the best the company has made in terms of the returns. We make vast investments in basic research as we have the resources to do that. In some sense, there are many businesses Microsoft is able to be in because of Microsoft Research and that’s the statement about the impact. Microsoft Research, in some points of time, is critical in our ability to be competitive.
Across your five Microsoft Research centres in Washington (Redmond), California (Silicon Valley), England (Cambridge), China (Beijing) and Bangalore, what are the projects that excite you the most?
There are a lot of things that excite me. I have been in the computer science field for more than 30 years and if I look at the research that’s been done around the world right now in many areas, I see more progress than I have really seen in a long time against a number of really critical problems that people didn’t think you could solve. The work we are doing in computer vision is changing the way people think about images and how they use it, including something called HD-view which is being able to take thousands of photographs and the technology uses computer vision to match up all these pictures and make them look like they have been taken at the same time even though the lighting may have been different, and remove things that are inconsistent in the pictures. We have also moved into the era of data.
Going forward what’s going to be the next big thing in computer sciences?
From a technical perspective, the biggest thing is that our notion of what computers are, is under tremendous amount of evolution right now as we move to these highly parallel systems such as the multi-core and mini-core kind. Another change is that we are building the most enormous data centres that have ever been created. Microsoft, Google, Yahoo, Amazon and eBay constitute 10-20% of the entire personal computer (PC) server business. The real question is that once we build these things then what’s going to happen. The PC became popular to do one thing and then all these other applications came about and there was an explosion of creativity. What we are seeing here is similar in that these large-scale data centres and data systems are being built for a well understood value and once they exist, new opportunities show up.
Are there any particular verticals that your research will focus on in the future?
Microsoft has turned to the health-care business. The opportunity to make a tremendous change in medicine by being able to do really large-scale analysis of medical data is one of the areas where we will see tremendous change in the next few areas and these large data centres enable us to think about that. I don’t know if that’s going to be the next big thing or not but it is a big change and will have a huge impact.