Technology, for the sake of it, is not useful in the long run, says ARM COO
Bangalore, June 18
Designing chips is challenging work - designers try to keep up with shrinking chip geometries and the fast-paced consumer electronics market.
Business Linecaught up with Tudor Brown, Chief Operating Officer of ARM, who was in the city inaugurating a new office, and questioned him about the tech landscape. ARM is a chip design firm based in the US.
What are the recent trends you have observed in India?
The job market has tightened over the last 18 months. It's not easy to find talent in Bangalore, and we are now recruiting from Delhi, Chennai, etc. This is an indication of a competitive market, but ARM needs to start partnering with institutes and develop talent. Those days are over when an ad was enough to get people. We need to take a longer-term view and start working with colleges here. (The company plans to begin partnering with institutes of higher learning and engage PhDs for research work soon.)
How about global trends - what are the challenges arising in chip design?The microcontroller (a microcontroller is a chip that includes microprocessor, memory, and interfaces) story for ARM has not yet begun. We believe sales of microcontrollers will climb once the hardware cost is addressed. Prices are set to drop and in 2010, you can expect the price of a 32-bit microcontroller to be equal to that of an 8-bit microcontroller. It's currently five times the price. However, software will remain a barrier.
In what way?Simplistically, everything is getting harder. Process technologies are moving from 90 nm to 65 nm and 45 nm. But the tools are not advancing as fast. The tools to verify and ensure you get the design right are not mature enough right now. This needs to improve. It's time to implement a `parallel paths' methodology. In this model, hardware and software are developed in parallel and in co-ordination with each other.
How is ARM keeping up with these asynchronous industries?We are talking to OEMs, in an initiative called `Segment Marketing'. We track trends in the consumer electronics market and also keep in touch with OEMs on what they are working on, so that we can design better chips that help their gadgets function better. We have learnt a lot in this manner - one example is when we talked to a hard disk maker, and added DSP (digital signal processing) to the chip that controls the arm of the reader. This, unexpectedly, opened up an entirely new market for ARM in MP3 decoders.
How are open source operating systems such as Inferno and Linux doing in the chip architecture domain?Currently, there is some uncertainty in the industry over the use of open source. Firms view it as an intellectual property (IP) liability. This is primarily because of various lawsuits, and open source's maverick image. This will change, though, once people understand there's no such thing as `free IP'.
What is ARM working on in research and development?Performance monitors on chips, similar to the ones that offer PC users a graph on the performance of their machines, is new. However, this will be hidden to the user and is being built only for more efficient internal functioning of the chip. Other themes are more efficient designs - for higher performance; the intelligent energy manager that guarantees better battery life, so even fancy phones last longer; reducing costs of design; and embedded security for when the mobile will become your credit card.
As complexity increases and costs decrease, what can the layman expect to see in the future?More features in gadgets, at the same price. The consumer electronics industry can expect to see more complex products at lower cost. Taking an example of a minimalistic MP3 player used nowadays, it will be replaced with a high-end one that boasts a display, multiple codec standards (so different formats of songs/videos can be played) and longer battery life. You see, technology, for the sake of it, is not useful in the long run.