You cannot limit Intel CTO Justin Rattner to chip chat. He heads the labs of the world?s largest microprocessor company, but has enough neat stuff to cause major disruptions in consumer electronics industry. Here to introduce an experimental 48-core single-chip cloud computer (SCC) processor in Bangalore, he also met ministry of human resource development officials in Delhi to give them insights into the latest educational technologies. Pragati Verma caught up with Rattner to understand the disruptive work Intel is currently championing in the world of computing and consumer electronics, including aggressive multicore designs, research to put data centres on a chip and 3D TV. Excerpts:
You have just demonstrated an experimental 48-core processor that has the potential to reshape computers and other gadgets? Do you expect entire data centres to be implemented in silicon?
Many of these services that we have been talking about will rely on the enormous power of computing in the cloud. This means we need enormous server capability and enormous storage capability in the internet. But the well known online companies have been taking enterprise computing solutions and datacentre solutions to build much bigger consumer datacentres also called IP datacentres. These are not energy optimal.
The experimental processor that we call 48-core single chip cloud computing is intended to explore alternative ways of delivering maximum computing power in a much more energy efficient fashion way. It will also benefit enterprise computing because it will deliver computing to hundreds of millions of users through virtual enterprise datacentre. That caused a complete rethinking on how you create an application as common as email. What we are trying to do with this single computer on a chip (SSC) is to understand what the requirements are for this consumer datacentres or IP datacentres and how to bring down the cost through cloud based computing.
When do you expect it to be out in the market?
We have successfully built the research prototype and in few years, we will see products in the market. I think we are looking at 2013-2014. We have got several years of work to do on the research side before we can commit to product developments and the feedback from our research partners like Microsoft is going to be very important. It was such a big problem that we decided to engage the entire cloud computing community to conduct the research.
This is probably the first time Intel undertook this kind of an experiment. What was Indian lab?s contribution?
Yes, this is the first time we built anything with the intent of making it available to the research community. We have a Silicon prototype team in Bangalore. This team specialises in creating experimental, high value and complex microprocessors. They did the 80-core processors few years ago and were very successful. This is dramatically more complex?more than 10 times complex. Their contribution was very much in the silicon design. We also have a silicon team in Hillsboro, US and software evaluation team in Germany. These three teams worked for about 18 months to develop the processor and about 40 people worked on it. That?s probably one-twentieth of the number of people working on the production class processors. In India, we are also doing research on high security systems that could be used in financial transactions. Then, we are working on a kind of hybrid between storage and security. We also do some research on health related technologies.
You also met ministry of human resources officials. Are you planning a fresh education initiative with the government?
We talked about the role technology plays in education, both in basic and higher education. We have been very active in India with our Intel Teach programme. We train million and a half teachers roughly on the use of technology in the classroom. We also talked about some future technologies that can play a major role in education. We saw a lot of interest in virtual world, 3D environment, where teachers can teach science, mathematics and the other subjects. Also, there is a lot of interest on TV 2.0, the future television which is more user driven. You might think of it as a broadcast medium, but here the users can choose the type of information and the kind of content. They can browse and have access to TV content whenever they want, wherever they are without really having any dependence on broadcast schedules. It is not limited to certain formats and might be playable on the personal devices. Actually, I was surprised that there was so much interest in television from an educational perspective. At Intel, we are working on the research side and starting to work with some of the product appreciation and these are the things we can look out in the next 3 to 5 years. For example, 3D will come out very quickly in the market in a year or two. Three years from now, every TV will be manufactured with 3D capability, which will drive the cost down. So, you can have a 3D display at your school, if not individual classrooms.
Your vision of a personalised TV seems to suggest that we will have television sets with an ?Intel Inside? soon?
Absolutely. In fact, we have been focused how to deliver the 3D experience. We have a business unit called Digital Enterprise and they are building a whole family of products and solutions for digital television. There will be a few products which will be shown in the Consumers Electronic Show in Las Vegas in a couple of weeks including set-top boxes, satellite boxes and digital video recorders. I think Intel will be inside many of these kinds of devices in the coming years. It is pretty interesting that we are thinking of taking our Intel architecture to the consumer electronics space. At the same time, consumer electronics people want to integrate internet experience into television and there is nothing better than Intel architecture for this. And for us, we are in the right place at the right time as the consumer electronics industry is pretty excited about bringing Intel on their platform.
I am a little scared that if TV becomes another PC?
We appreciate your concerns. It?s not going to be a PC for sure. I think television is a very different experience and we spend a lot of time with folks like Yahoo and others on creating a user experience around television which is quite different from PC. The experience is to bring internet experience without disturbing television experience. Intel is playing a major role in the change of product design in the television space by understanding the user experience in a very deep sense.
What did your study show? What are the biggest needs of a TV user which you think are not being fulfilled today?
I am not an expert here. But I would think technical people and even consumers don?t want internet experience as the browser experience. Regular PC browsers like Explorer, Chrome or Firefox are not workable in TV sets. So the challenge is to create, use and present internet content when the screen is 10 feet away and you don?t have the ability to click a URL with a mouse. So, how will you make internet relevant for television without using the traditional browser on the screen. We are working on widgets for the TV sets which can provide similar internet experience. Widgets are not like browsers, it?s just an app.
Which screen do you expect to become most popular?TV, mobile phone or PC?
I don?t think there will be one favourite screen. We are talking about accessing information and content whenever I want and wherever I am. If I want it on a small screen, I should get it on a small screen. The way to access information should suit the screen size. Be it cellphones, desktop, netbook, or even television, each of them has a different way of accessing information. If you want to watch a blockbuster movie, you would like to watch it on the biggest screen available to you. You don?t want to watch Titanic on a cellphone; you will not have the right experience. So, challenge is for product designers to move across these different screen sizes and yet offer the same user experience.