Interview with Hariharan Subramanian

Featured Engineer

Hariharan Subramanian

Hariharan Subramanian - Graduate Student, Viterbi School of Engineering, University of Southern California

What are your favorite hardware tools that you use?

I predominantly use Cadence suite for my IC and ASIC design adventures, ModelSim for preliminary simulation of my HDLs prior to synthesis. I also believe that Ocean script makes Spectre very powerful, that makes it my currently favorite simulator.

What other tools help your design process?

Matlab and Mathematica are two very useful tools, with the former gaining more importance for any mixed-signal design (for emulating the DSP blocks). The importance of MS-Excel for IC Design can never be put on words!

What is the hardest/trickiest bug you have ever fixed?

There have been several instances where technical faults have caused my robots sit idle; processors telling me 2+2 is not 4! This one incident happened when I was building my Robonoid, a multi-tasking robot. I was not able to understand why the motors weren’t moving while after sometime I felt the controller used is heating up. The motor started showing some movements but at short bursts! And finally a puff of smoke indicated that there was a spare wire that accidentally shorted the supply rails! It turned out that I had to replace my batteries, regulators and the PCB to bring my robot back to life. I can understand the faults in my circuit or HDL code and debug them but a misplaced wire under the PCB? A hard one though!

What has been your favorite project?

I have considered all of my projects as a mission with some objective rather than a task! With each project I reinvent myself by improving my work techniques, thus able to evolve with more new projects time and again. My research project, ‘i-Bran: Electronic Braille Enterprise’ is my favorite project. Each and every step, from identifying the motive to filing a patent for it, it was a great experience, given the multiple dimensions of the project and the extremely realistic constraints that provided a very little to no margin of error.

Can you tell us a little more about your project i-Bran?

It is a portable device that can be used by visually impaired individuals to read books. Books, in the format of .txt files, stored in CD-ROM or SD memory card can be directly used in the i-Bran. The small size helps the blind people to carry it wherever they want to. Another important aspect of i-Bran is its interface with computers. It can be connected with any PC at the touch of a button.

I came up with this idea basically to provide a practically feasible support for the children of blind school to browse www. I designed this project and implemented it with my project team members under the guidance of Dr. S. Rajapandian, PhD, IISC. I am happy to say that out of several hundred projects i-Bran was specifically chosen by my undergraduate school and awarded full funding.

From the beginning we were constantly in touch with a Blind School. We held several discussions with the school HeadMistress and several of the school students to determine what are the prime features to be incorporated in our i-Bran. Each and every feature in i-Bran is meant to solve some particular hardship faced by the school children of that school. Most of all, we had a very simple motto: No Compromise.

Another reason for this being my favorite project is that we introduced a new method of teaching that can be followed in blind schools. This reduces the burden of using Braille text-books. The device which we designed for a cost of approximately $50 can be used by every student in the class with one PC controlling them. The PC shall be handled by the instructor.

Though this project on the whole is an invention, it included several finer aspects all of which were a first, worldwide. For instance the scrolling mode of Braille display is one that the students were very happy about. We developed three versions of i-Bran with each version showing added features designed aesthetically for the students. I-Bran v1.0 will remain our personal favorite!

I must say that the day we tested the i-Bran at the blind school, it was a very proud and emotional moment for all of us when a student read out the alphabets slowly from our device.

Can you explain about your transition from embedded projects to IC design and Processor design projects?

It is not that I did not like Circuit Design. In fact for i-Bran & for all of my Robotic ventures I designed each and every circuit, including soldering of PCB! I personally feel that a great deal of work is still left to achieve a ‘sustainable technological development’. Restricting my knowledge to microcontrollers is not good for my interests. In future I would like to be involved in the development of a product that will impact the everyday life of a common man and will push the hardware industry forward. Digital and Analog system design is my field of interest. I can personally see how my research and project experience in the various fields of Computer Architecture, IC Design and Hardware validation aid me in bringing out the best product possible.

I am also involved as a Research student at the Viterbi School of Engineering in the Retinal Neural Circuits design. This is one research that will break open a lot of barriers.

Do you have any tricks up your sleeve? (Special way to analyze circuits, special process you use to make something, etc.)

Intuition! For a hardware engineer the knowledge about the field is helpful in designing the product. But it is the intuitive thought process at each stage of the design that makes the product better than the others. Rather than what you are positively trying to do, think also about what the circuit should not do. As an example, the way I test my designs involve extensive test cases and rigorous test runs. But surprisingly when a fault occurs I tend to spot it very quickly because of the way I design; I share a sixth sense connection with my designs that will provide me with specific information which can be used as a map to the faults in the circuit.

What is on your bookshelf?

As a hardware designer I always have the John L. Hennessy & David A. Patterson ‘Computer Organization and Design’, CMOS VLSI Design by Neil Weste and David Harris, and ‘Design of Analog CMOS Integrated Circuits’ by Behzad Razavi. But for specialized topics I refer to a number of books, for instance, Digital Signal Processing by Alan V. Oppenheim or Data Converters by Franco Maloberti when it comes to my data-converter project. The former list is relatively constant compared to the latter.