I chose to study computer science at the University of Sydney primarily because of the excellent computer systems group led by Professor Zwaenepoel. I wanted to complete my degree under his guidance and I was excited about experiencing a new academic environment. The University of Sydney provided a really friendly and open environment, where I felt supported to conduct my best research. It was also a great experience to spend a few years in a city as vibrant as Sydney.
My interest in computing started in high-school in Romania, where I took my first programming class. Even though I found it challenging, I remember the great satisfaction I felt when my first (very simple) programs would run correctly! After that, I was pretty much hooked. I continued this path through my undergraduate studies at EPFL in Switzerland and the journey ultimately brought me to Sydney for the PhD. I think that my parents’ careers also influenced my appreciation for computer science - my mother is a computer scientist and my father is an electrical engineer.
My PhD research focused on the design of efficient key-value stores, which have become the standard storage platform for many cloud applications. The kinds of systems I studied in my dissertation are backing the storage infrastructure of companies such as Amazon, Google, Microsoft, and Apple. I was drawn to this research area because it provides many intellectually stimulating problems, which are also practical. I find this combination very motivating.
I developed several techniques to improve the performance of storage systems, such as making them scale with the degree of multi-threading, taking advantage of large memory sizes, and decreasing the impact of system maintenance on user operations. My designs are conscious of the properties of new storage hardware and thus lead to large performance improvements in terms of throughput and tail latency. This work is practical and can be applied right away to data-centre storage systems. In fact, part of my work is already deployed in industry at Nutanix, a cloud computing company that I have been collaborating with throughout my PhD.
I felt really happy and honoured that my dissertation was chosen as the winner. In addition to the sense of fulfillment coming from my work being recognized by the community, the visibility generated by winning this award is really important in attracting talented students as a junior academic at McGill.
One thing the PhD taught me is that every student experience is different and thus support looks different at each stage of the student journey. I will try to understand the particular situation of each student when advising and teaching and do my best to provide them with the resources that they need to succeed.