Microscopic quantum mechanical interactions among heat carriers called phonons govern the macroscopic thermal properties of semiconducting and electrically insulating crystalline solids, which find applications in thermal management of electronics, thermal barrier coatings and thermoelectric modules. In this talk, I will first describe my recent work on how our newly developed first-principles computational framework to predict these microscopic interactions among phonons unveils a new paradigm for heat conduction in several of these materials. As an example, I will describe a curious case of heat conduction in the ultrahigh thermal conductivity material - boron arsenide (BAs), where the lowest order interactions involving three phonons are unusually weak and higher-order scattering among four phonons affects the thermal conductivity significantly, in stark contrast with other ultrahigh thermal conductivity materials like diamond and boron nitride [1, 2]. I will show that this competition between three and four phonon scattering can be exquisitely tuned with the application of hydrostatic pressure, resulting in an unusual non-monotonic pressure dependence of the thermal conductivity in BAs unlike in most other materials . Finally, I will describe my prior experimental effort to probe the scattering of THz-frequency thermal phonons at atomically rough surfaces of a nanoscale silicon film using a non-contact optical pump-probe experiment called the transient grating. Through these measurements, I will show that the surface scattering behavior of the heat-carrying thermal phonons is extremely sensitive to the changes in surface roughness of just a few atomic planes even at room temperature .
References [*: Equal Contribution]:
 Fei Tian, Bai Song, Xi Chen, Navaneetha K. Ravichandran et al., Science 361 (6402), 582-585, 2018
 Ke Chen*, Bai Song*, Navaneetha K. Ravichandran* et al., Science 367 (6477), 555-559, 2020
 Navaneetha K. Ravichandran & David Broido, Nature Communications 10 (827), 2019
 Navaneetha K. Ravichandran, Hang Zhang & Austin Minnich, Physical Review X 8 (4), 041004, 2018
With increase in usage of the Internet, there has been an exponential increase in the use of online social media on the Internet. Websites like Facebook, Instagram, Twitter, Tinder, YouTube, and many others have changed the way the Internet is being used. There is a dire need to investigate, measure, and understand online social media from various perspectives (computational, cultural, psychological). Real world scalable systems need to be built to detect and defend security and privacy issues on online social media. I will describe some cool projects that we work on: Distracted Driving, Elections, (Mis)Information / Fake content on Twitter, Facebook, and WhatsApp; Selfie deaths and KillFies; and Manipulated followers. Many of our research work is made available for public use through tools or online services at precog.iiitd.edu.in including datasets http://precog.iiitd.edu.in/resources.html that we have used in our research work. Our work derives techniques from Computational Social Science, Data Science, Statistics, and Network Science.
The rise of multi-core systems has necessitated the need for concurrent programming. However, developing correct, efficient concurrent programs is notoriously difficult. Software Transactional Memory Systems (STMs) are a convenient programming interface for a programmer to access shared memory without worrying about concurrency issues. Another advantage of STMs is that they facilitate compositionality of concurrent programs with great ease. Different concurrent operations that need tobe composed to form a single atomic unit is achieved by encapsulating them in a single transaction.
Most of the STMs proposed in the literature are based on read/write primitive operations on memory buffers. We denote them as Read-Write STMs or RWSTMs. Some read/write primitives result inunnecessary aborts. To ignore unimportant lower-level read/write conflicts and allow better concurrency,we design and implement efficient Object-based STMs (or OSTMs) with concurrent closed addressed hashtable (HT-OSTM) and list (list-OSTM) which export the higher-level operations such as insert, delete,lookup, etc. as transaction interface.
It was observed in databases and RWSTMs that storing multiple versions provide greater concurrency. We combine both of these ideas carefully for harnessing greater concurrency in STMs: multiple versions with objects semantics. We propose the notion of Multi-version Object STMs or MVOSTMs. Specifically, we design and implement MVOSTM for the hash-table and list, denoted asHT-MVOSTM and list-MVOSTM. We prove the correctness of proposed MVOSTMs as opacity and show significant performance gain over state-of-the-art STMs through experimental evaluations.