Thursday, February 26, 2015

10 Questions with SC15 Student Programs Chair Jeanine Cook

In the months leading up to SC15, the conference communications team will be highlighting various aspects of the conference by asking questions of the volunteers leading those areas. 

In this first installment, SC15 Student Programs Chair, Jeanine Cook, talks about plans to broaden the size and scope of student activities at the conference.
Jeanine Cook, SC15 Student Programs Chair

Cook is a principal member of the technical staff at Sandia National Laboratories in Albuquerque, where she works on several projects aimed at paving the way to exascale computing. Prior to joining Sandia, she was an associate professor in the Klipsch School of Electrical and Computer Engineering at New Mexico State University.

She earned her B.S. degree in Electrical Engineering from the University of Colorado, Colorado Springs, in 1987, an M.S. degree in Computer Science from the University of Colorado, Boulder, in 1996, and her Ph.D. from New Mexico State University in 2002.

In 2008, she was recognized with the 2008 Presidential Early Career Award for her research in computer processor performance modeling. Another distinction not on her CV is her being “awarded” a ticket for speeding…in her wheelchair…in the Austin Convention Center at the SC08 conference. She clearly takes the idea of high performance seriously, no matter what she is pursuing. And when she wants to get away from it all, she’ll saddle up her horse and ride out into the New Mexico desert or wilderness. She enjoys reading about the history of the west as much as she likes exploring it. 

At SC15, most of the programs developed for students will be under a new umbrella – Student Programs. Can you tell us what that’s about? 
Essentially what we want to do is build a bigger student community at the SC conference, rather than having the students pocketed here and there, like in Student Volunteers and Experiencing HPC for Undergraduates. So we captured the pieces of the conference that students have been a part of to create a general student community. Then we plan to take that whole student community and plant it into the larger community. We’d like to make the student community more visible and encourage all attendees to mingle with each other. Ideally, at some session, a big group of students would show up and other attendees would wonder, “Who is that giant group?” We think if the students are more noticeable, they will become more engaged. 

What prompted this realignment?
A couple of things. Among the SC planning committee, we felt we weren’t doing a good enough job of integrating the students into the larger community. One of our concerns was about the next generation of HPC researchers – were there enough replacements coming up and how can we help them succeed? Another concern is that we want to make sure that students who attend SC remain engaged, that they submit posters and tech papers, as they continue in school and begin their careers.  
Students take on a variety of important roles at SC conferences
How will you know if this new approach is successful?   
We will submit a proposal to do a formal evaluation, hopefully over several years so we can continue to make improvements. We will also look for anecdotal data, such as asking students about their experiences. We also want to hear from other attendees about their experiences with the students, whether they served as a mentor, worked with a Student Volunteer, or whatever. 

You joined Sandia National Labs after 11 years as a professor at New Mexico State University. Based on that experience, how are you looking to make SC a richer experience for students? 
Well, I started out in industry as a rocket scientist, working for a division of McDonnell Douglas, so I come from a variety of different professional places. I hope we can let students know about the good parts and the not so good parts of various careers, so they can develop an idea of what to expect in each of those areas. We want to give them a balanced perspective. 
Jeanine Cook

You also received a Presidential Early Career Award in Science and Engineering in 2008. What do you see as key factors for succeeding in those critical early years of a person’s career in HPC? 
My answer is that networking is very important. I kind of hate to say that, but I think it’s a big reason why I was successful. When you agree to give a talk here, attend a meeting there and serve on a committee over there, you meet a heck of a lot of people. So networking is super-important, just as being good at what you do is very important.

When I agreed to serve on panels, it wasn’t because I wanted to review a lot of proposals. You want to get people to know who you are and what you do. Then when they are putting together a proposal and wonder “Who does performance analysis, they think ‘Jeanine Cook does that.’”

But growing a network can be a painful process for students and people just starting out in their careers. That’s why it’s important for us to integrate them into the larger community. 

Looking back, was there one person who played a crucial role in your success? 
I don’t think there is just one person I can point to. It was a lot of people, which goes back to my networking comments. I met a lot of program managers and academics. Early in my career, because I made an effort to get to know a lot of other academics, they would tell me what to do and what not to do. Other women were often my biggest advocates, especially in academia. 

Now back to the present…what are you working on now? 
Holy-scamoley guacamole. I’ve got too many projects! My big project now is to define a beyond-exascale architecture. We’re looking at a processor in memory and storage type of solution. It’s not cmos-based. The goal is to define both the architecture and the programming model. We think it will have applications in areas other than scientific computing, such as neuromorphic systems.

Another big thing is understanding performance tool requirements for next-generation systems, like exascale. There are currently no tools to adequately measure performance at the exascale, so we’re looking at existing tools, then determining what tools we’ll need for next-generation systems and how we’ll build them. I love architecture, but I really love performance modeling and analysis. I like having lots of data, plotting it and figuring out what’s happening.

My other big project is Sandia’s Structure Simulation Toolkit. We have a team of amazing people and we’re trying to develop a framework for large-scale simulation of next-gen systems. At New Mexico State, I worked with a lot of grad students to create models of processors and we developed pretty accurate techniques. Now we are re-engineering them for large-scale modeling. 
Finally, I’m working with CAL, the Computer Architecture Lab project for DOE that’s developing abstract machine models. And I’m working on a resilience project, looking at what fails on current machines so we can try to get a handle on what will fail on next-generation systems. 
Jeanine on her horse "Ringo Star".
Some of your colleagues have commented on your intense focus once you get rolling on a project. What do you do to get away from it all? 
Eat chocolate and drink Pinot Grigio. Actually, riding my horse is my number one thing to do for fun. He’s the most beautiful, cutest horse on the face of the Earth. You will find no cuter horse. He’s a Tennessee Walker and his name is Star’s Red Ringo, but we call him Ringo Star. The beautiful thing is that from our house, you can ride for 100 miles on public lands. If you stand in our driveway and do a 360, every mountain you can see, I’ve been to the top of. Riding in the desert is fun, but the mountains are even funner.

Sometimes we come across old graves or graveyards. One time we were at Fort Mason, which was a fort on the old Butterfield Trail, and my husband plopped me down near the old trash pit. I found some old nails and pieces of old plates, but I’d really like to find an old relic, like the 132-year-old Winchester rifle someone recently found resting against a tree in Utah. That would be so cool to find something like that.

One of Jeanine's favorite activities.
With SC15, the conference returns to Austin. When you were there for SC08, you got a speeding ticket inside the convention center. Would you care to explain that? 

No comment.

One final question. You’ve been known to ask people “Who’s your momma and where’d you come from?” How would you answer that? 
I’m a Michigan State University baby, born in Lansing where my dad was earning his Ph.D. in physics. He then got a job with the National Bureau of Standards and we moved to Maryland. He then got a job as a government contractor and we moved to Colorado. I loved Colorado until it got too crowded, so I ran away to New Mexico in 1996 and finished my Ph.D. I got a job here, my husband got a job here, and here we are.