My research interest lies at the intersection of data, ML and system. At Meta AI, I'm exploring limits of scaling for many SOTA models.

Jun Yang is a William Kepler Whiteford Professor of Electrical and Computer Engineering Department at the University of Pittsburgh. Prior to joining the University of Pittsburgh, she was an assistant professor of the Computer Science and Engineering Department at University of California, Riverside. Jun received her bachelor from Nanjing University, China, and her PhD from the University of Arizona in 1995 and 2002 respectively. Jun’s research is in the broad area of computer architecture and her recent focuses include GPU designs, architecture level security, emerging memory technologies, 3D integration, and power and thermal management techniques. Jun is a recipient of NSF CAREER award in 2008, IEEE MICRO Top Picks award in 2010, and best paper awards of ISLPED 2013 and ICCD 2007. She was on the editorial board of IEEE Computer Architecture Letters, and she has served in the Organizing and Technical Program Committee in ISCA, MICRO, and HPCA, for many years. She has been included in the HPCA hall of fame since 2017.

Thesis Title: LongLiveNoC: Wear Levelling, Write Reduction and Selective VC allocation for Long lasting Dark Silicon aware NoC Interconnects

Increasing processing demand has led to the development of chip multiprocessors which can have multiple to many cores connected with each other and with the on-chip caches. These connections are established by an on-chip packet-switched Network-on-Chip (NoC). Scaling of technology nodes increases the power dissipated by the chips leading to thermal restrictions. To control the chip thermal design power, certain components (like cores and caches) may be turned off. However, in this scenario of dark silicon, the interconnect is expected to be available.

The thesis aims to save power consumed by this always ON interconnect by replacing the power-hungry SRAM buffers in the routers with low leakage Non-Volatile Memory (NVM) based buffers. However, the major challenges with the employment of the NVMs are slower writes and weak write endurance.

The thesis proposes:
1. Methods to evenly distribute the writes across these NVM buffers in order to increase their lifetime. This is done by static and dynamic allocation of buffers to the virtual channels and their selection during packet transmission.

2. Power can also be saved by using frequency scaling of the routers and/or turning off certain buffers when the usage is less. The investigation is done for all such approaches and power savings are demonstrated.

3. Endurance can be improved and energy can be saved if we can reduce the number of writes performed on the buffers. This is achieved by proposing two compression techniques leading to reduced network traffic and improved lifetime.

All the above methods help in improving the lifetime of the NVM based NoC interconnects in the context of dark silicon.

I am a PhD candidate at the University of Illinois at Urbana-Champaign. My advisor is Prof Sarita Adve. My research interests include Error-Efficient Computing, Approximate Computing, Hardware Resiliency and Software Testing. The overarching theme of my dissertation work is enabling reliable, low-cost and efficient computing by allowing controlled errors in the system. My work aims to explore such opportunities in emerging workloads and build an ecosystem to formalize and automate the application of error-efficient computing techniques. I am among the young researchers selected to participate in the 2018 Heidelberg Laureate Forum.

Before joining UIUC, I was a CPU/Silicon validation engineer at Intel for five years, where I won a divisional award for key contributions in validating new industry standard CPU features. Prior to that I worked for 1.5 years at Qualcomm on the architectural verification of the Snapdragon processor. I completed my MS in ECE from North Carolina State University. My undergraduate degree, in Electrical Engineering, is from the University of Madras in India.