Irina Tezaur

Irina Tezaur helps study ice sheet dynamics, as represented by this projection of Antarctica

The road to Livermore has been long but rewarding for Irina Tezaur. Today she is an award-winning computer scientist at Sandia National Laboratories, but in 1992, when her parents emigrated from Russia, she was 8 and spoke no English.

Her parents, both PhD scientists, settled in what was then a Russian-speaking suburb of Detroit.

There, with the help of a sponsor, they found work, but nothing commensurate with their educations.

Her computer scientist mother got a basic programming job – not exactly challenging for someone with her background, but at least it was in her professional field. For a time her father, a physicist in the field of optics, had to work as a janitor.

In the international community where they settled, Tezaur went to school with other children who did not speak English, schools that employed tutors in languages like Arabic and Russian.

As difficult as those early days might have been, Tezaur looks back at her family’s experience in the U.S. as something of an American Dream.

Within a few years, Tezaur’s parents found more rewarding jobs and were able to stretch their finances to buy a 2,000-square-foot house in a nearby town.

She herself graduated from an Ivy League school, earned a PhD at Stanford and now is a computer scientist at the Livermore campus of Sandia National Laboratory.

Her abilities and contributions were recognized this summer when she was given a prestigious Presidential Early Career Award for helping find ways to make complex computer programs run faster and more efficiently.

Computer modeling is the heart and soul of much of today’s science, but programs can be prohibitively time-consuming and expensive to run as investigators try to understand nature in ever-finer detail. As fast as today’s supercomputers are, some of the biggest programs can take weeks or even months to run.

In a recent article in the Sandia Lab News, Tezaur described the challenge of trying to coax computer models to simulate large and tiny objects at the same time.

“Models are really hard to build when there are different scales involved – that is, when you have an object that has really small pieces together with really large pieces,” she said.

“If you use the smallest scale everywhere to ensure that your simulation is able to resolve everything, you will end up with a really crazy amount of computations.”

She and mechanical engineer Alejandro Mota found a solution by extrapolating from methods developed by 19th century German mathematician Hermann Schwarz for dealing simultaneously with two different kinds of geometry.

Her efforts at streamlining computational processes have contributed to Sandia’s core nuclear weapons mission but also to climate research.

Since 2012, she has been a lead developer on the land-ice component of a U.S. Department of Energy climate model, the Energy Exascale Earth System Model. She has helped improve simulations of ice sheet dynamics that are essential to predicting future sea level rise.

Not surprisingly for the daughter of two scientists, Tezaur showed early proficiency in mathematics as a secondary school student.

Urged by her parents, “It was never really an option for me not to do well in that area,” she said.

For college, she went to the University of Pennsylvania, where she majored in mathematics. This was “pure” math as emphasized by the Penn math department at that time. Her advisor was a Russian mathematician interested in fractals.

At the same time, she wanted to experience math applied to practical problems, so she took actuarial classes through Penn’s Wharton Business School. While she learned that an actuarial career was not for her, she also concluded that she didn’t want to stay in pure mathematics. She would gravitate toward something that she now calls “impactful,” perhaps research that could prove to benefit society.

At Penn she did well enough to have choice between two prestigious graduate schools, Princeton and Stanford. She visited both and decided on Stanford, in part because students there “seemed much happier.”

Perhaps more importantly, the Stanford program, known as the Institute for Computational and Mathematical Engineering, was flexible and interdisciplinary.

“I would be able to work with people in many departments. I didn’t know what I wanted to do at the time and I thought having more options would be a positive.”

As it turned out, one of her Stanford professors had a consulting contract with Sandia’s main branch in Albuquerque. In time, this connection opened a summer internship opportunity in the branch of computer science in which Tezaur has made her contributions.

The Sandia project was called computer model reduction -- an effort, as she describes it, to “use fewer resources to solve the same modeling program.”

As a Sandia intern from 2007 to 2011, she found that she liked not only the work but the location, and the people she interacted with. She found herself ever more attracted to research, especially on this kind of “impactful problem that was of interest to national security.”

She became a full-time employee at the Albuquerque site in 2011, moving to the Livermore site in 2014. She married early the following year. She and her husband, a senior research engineer at Stanford who also does computational modeling, live in Union City.

Today, she still visits the Albuquerque site from time to time because so much of the Laboratory’s model streamlining work goes on there. On the other hand, communication programs like Skype make it convenient and economical to work productively at a distance and reduce the need for travel.

From national defense projects to climate research, she finds the collaborative atmosphere at Sandia helpful and creative. She is quick to give credit to others.

“For these large, high-impact projects, you need people from a lot of different areas coming together and working together because no one is going to know everything.

“These problems are very complicated.”

She appreciates the “ton of resources” available at the laboratory – “way more than I ever had in academia, in terms of computing platforms, access to supercomputers including new up-and-coming architectures.”

On the other hand, “I think the biggest resource is the people. There are a lot of very smart people who have a lot of expertise in different areas. They are very open to sharing, to collaborate and explore new ideas.

“The things I have received recognition for – it’s far from just my own work. It would not have been possible without collaborations and without all these people who have ‘been there’ to talk to and brainstorm with, and who have supported me.”