I'm currently an Assistant Professor at the CSIS Dept. of BITS Pilani, Goa Campus where I also lead the "Confluences of Computing" Lab (Confluence Lab, in short), emphasizing the confluence of science, technology and society and therein the role of computing.
My overarching interest, and a key objective of the Confluence Lab, is in building computer systems, right from atomic all the way to planetary scale, such that they are exceedingly scalable, intelligent, dependable, efficient and usable. We explore various paths to achieve this through novel computing and programming paradigms.
Academic Experience:
- Technion, Israel - Postdoctoral Fellowship
- UC San Diego, USA- PhD, CSE Department (May 2019)
- IIT Madras, India - MS - By Research, CSE Department
Industry Experience:
- Intel Labs
- AMD Research
- IBM Research
My CV:
Contact Details:
Official Email Id: kunalk [at] goa.bits-pilani.ac.in
Office Phone No: +91-832-258-0810
Office Room No: D-246
Expertise and Interests:
I like to deal with computer system design and implementation problems spanning software or algorithms all the way to hardware or architectures.
Typically, the type of problem solving that I do, involve learning the following four ropes:
- First, understanding relevant mathematics and computer science theory to address the problem at hand,
- At the same time, devising suitable programming languages and models for the target system,
- Mastering the target application domain and the algorithms of that domain, and
- Finally, by considering any practical physical limits of the underlying circuit and device technologies.
I believe the above skillsets are essential for designing and building novel and useful computer architectures and systems of the future.
Apart from core computer systems, my research work spans diverse topics in computer science and engineering such as programming languages and models, distributed systems, databases and data management, operating systems, trust-centric computing, embedded systems, electronic design automation, etc. I see many of these topics through the lens of system design.
Other than 'Data-centric' and 'Scalability-centric' system design approaches, numerous other approaches that I undertake in my research and teaching are: 'Correctness first', 'Trust-centric', 'Energy efficient', 'Software-friendly hardware', 'Hardware-friendly software', 'End-programmer focused', 'End-user focused', 'Bio-inspired', 'Complexity-effective', 'Domain driven', 'End-to-end', 'Cost conscious', etc.
I also work on emerging computing and programming paradigms such as in-memory computing, quantum computing, neuromorphic computing, and such. Here again, I am interested in full-system design and implementation using these upcoming technologies. I believe, these technologies warrant and permit rethinking computer architecture and systems from scratch!
Research Directions:
Some of my current topics of interest and research are as follows:
1) Computer Systems, i.e. Software and Hardware and Their Combinations: In core computer systems, the current concerns are heterogeneous computing, specialization-generalization tradeoffs and interplay of software and hardware architectures. To what extent architectures will be 'algorithm-driven' versus 'data-driven' will be one of the central questions in future computer systems. Emerging technologies and associated algorithms in quantum computing, machine learning, neuromorphic computing, etc. will likely determine the outcome heaviliy. Past issues, such as parallelism, locality, efficiency, etc, have to be revisited in the wake of this question.
2) Data-Centric Systems and Associated Novel Computing Abstractions: In here, the idea is to make processing data-enriched by exploiting the scale, site or location, semantics or structure of data. I believe that complexity, manageability or just pure scale limits the design of current computer systems and therefore addressing scalability related issues in system design will become increasingly important. Right balance have to be struck between the use of edge systems (e.g. edge AI, edge data centers, edge networks) and centralized systems (large core data centers, public clouds, etc). Finally, I believe for computer systems to be a part of our long-term civilizational story, they need to become more dependable (i.e. trusthworthy, safe, secure, etc.).
3) Democratization of Computing with a Focus On Societal Issues: Human-centric, society-centric and even ecology-centric computers is the focus here. Under this topic, the attempt is to look at the larger needs of society, such as towards education, health, banking and other services, and our planet, such as towards sustainability, energy, etc. One of the goals here is to best enable a positive role of computing in these critical societal domains/areas. Solutions to the issues are intimately linked to the future of computing (points 1 and 2)
To Prospective Students: