Fundamental limits in computing:
The overarching goal of our research program is to understand how to design and fabricate scaled, compatible electronic devices and systems that can operate close to the fundamental physical limits of computing and even overcome those. Having driven the information revolution over the last couple of decades, CMOS scaling has come to end. Yet, there is an enormous room for improvement in the today’s computing substrates. This is because there is a significant gap between the performance of the state-of-the-art and the fundamental limits of computing. The theoretical energy minimum for an irreversible logic or memory operation is of the order of kT; today’s transistor technology dissipates six to seven orders of magnitude larger energy at the functional unit level. Such physical limits originate due to considerations related to thermodynamics, statistical mechanics, quantum mechanics and electromagnetism. Examples of these limits include the Boltzmann limit, the Landauer limit, the von Neumann limit and so on. Building upon the foundation laid by the likes of von Neumann, Landauer, Bennett and Feynman, we bring in the new physics and interesting phenomena in emerging and quantum materials such as ferroelectrics, antiferroelectrics, and strongly correlated systems to realize novel-yet-practical-and-compatible nanoelectronic devices that achieve the computational performance at par with the fundamental physical limits. The effects of such nanoelectronic computing devices ripple through the entire hierarchy of computing–we collaborate with circuit designers, computer architects, and algorithm experts to assess the impact of our physics-materials-device level work at the full system level.
Relevant Publications: Tasneem et al. DRC (2018). Khan et al. Nature Mater. 14, 182 (2015); Khan et al. IEDM (2011); Khan et al. Appl. Phys. Lett. 99, 113501 (2011).
Foundational work: Landauer. IBM J. Res. Dev. 5.3, 183 (1961); Likharev, Int’l J. Theoretical Phys. 21, 311 (1982); Bennette, Int’l J. Theoretical Phys. 21, 905 (1982).
Emerging materials for computing:
We utilize certain classes of materials that are characterized by distinct phase transitions and order parameters. Examples of such materials include ferroelectrics, antiferroelectrics, and strongly correlated systems. Our focus is to engineer long-range interactions & correlated phenomena in these materials to bring in new physics and functionalities in nanoelectronic devices. This caters to our overall goal to achieve device operation at the fundamental limits of computing as well as to realize non-CMOS like computing capabilities. Our material work involves growth, integration and detailed characterization.
Relevant Publications: Wang et al. Appl. Phys. Lett. 112, 222902 (2018); Khan et al. Nano Lett. 15, 2229 (2015); Khan et al. Appl. Phys. Lett. 105, 022903 (2014).
Primitives for brain-inspired computing, AI cores and non-von Neumann architectures:
Big data applications driven by machine learning and artificial intelligence underpin the current paradigm of computing. Yet, handling these massive data sets in conventional von Neumann computers leads to significant loss of performance and increase in power dissipation to shuttle the data between the logic and different levels in the memory hierarchy. The future of computing requires new types of non-von Neumann systems that can co-locate compute and memory operations in the same/nearby functional units as well as execute different unconventional functions in a compact efficient manner. Compared to the conventional CMOS technology, emerging nanodevices intrinsically combine logic and memory, and can lead to more energy-efficient and higher performance hardware platforms for such computing applications. Furthermore, unlike CMOS transistors, such nano-devices can map the dynamics of neuron and synapses in biological brains thereby leading to the possibility of extremely, efficient realization of brain-like computing cores. Our goal is to understand how to leverage the interesting properties of our nanodevices (such as stochasticity, hysteresis, polymorphism, non-volatility) for unconventional computing architectures. We heavily collaborate with research group working on circuits, architectures, systems and algorthms to connect our work at device level to the higher levels of abstractions.
Relevant Publication: Wang et al. IEEE Electron Dev. Lett. 38, 1614 (2017). Presentation: ISVLSI2018