Sandia’s Siva Rajamanickam
VANGUARDS OF HPC-AI: Siva Rajamanickam became interested in HPC in 2010 when he joined Sandia National Laboratories after earning degrees in computer science and engineering. As an individual contributor at Sandia he initially focused on scalable algorithms for HPC use cases, such as linear solvers, sparse linear algebra and graph algorithms. In 2016, he became a team leader for co-designing algorithms for AI-focused architectures.
Siva is recognized as an HPC-AI Vanguard for demonstrating qualities of exemplary team leadership. Over the past several years, he has racked up numerous achievement awards, establishing himself as one of the HPC-AI community’s current and future leaders.
An Interview with Sandia’s Siva Rajamanickam, Distinguished Member of Technical Staff, Sandia Labs
What is your passion related to your career path?
My passion is to enable new scientific discoveries using AI and high performance computing. I am passionate about working in the broad areas of machine learning for science, codesign of algorithms and architectures, performance portable algorithms, and combinatorial scientific computing. I focus on improvements in each one of these areas to enable new science and the mission of a laboratory that plays a critical role in the nation’s security.
Do you prefer working as an individual contributor or a team leader?
I prefer working as a team leader. I can see my impact has a large multiplier when I lead a large team to achieve our collective vision to solve a challenging problem. I also feel it is a way to give back to my scientific community when I lead the next generation of HPC and AI scientists and help them grow.
Who or what has influenced you the most to help you advance your career path in this advanced computing community?
I am fortunate to have some fantastic mentors who have directly and indirectly helped me grow. I have learnt a lot from leaders such as Mike Heroux, Jack Dongarra, and Bruce Hendrickson.
What are your thoughts on how we, the nation, build a stronger and deeper pipeline of talented and passionate HPC and AI professionals?
I believe US has a very strong and deep pipeline of HPC and AI professionals. At this moment, the demand might be outpacing the supply, especially in the field of AI. However, I strongly believe as a nation we can meet the demand by using technology, relying on our fantastic universities and by training our current workforce. I believe we live in a world where adapting to change is necessary and somewhat easier now.
What does it take to be an effective leader in HPC and AI?
I think it is the same as what it takes to be an effective leader in any other field that changes rapidly. It requires an equal parts ability to adapt to changes and a strong conviction to focus on your long term goals. The rapid changes in the convergence of HPC and AI also require multi-faceted leaders with a broader view of many fields and who at the same time maintain deep knowledge of several areas and technologies. Finally, one cannot emphasize more the importance of being authentic and the ability to keep a team of entrepreneurial individuals focused on a common goal.
What is the biggest challenge you face in your current role?
The biggest challenge we face is to co-evolve algorithms and architectures so we can build a scalable way to train future AI models on HPC systems. I believe the current path we are on is not sustainable and we need new approaches that are scalable and robust. The second biggest challenge I see is a way to integrate several co-pilots in a productive manner.
What changes do you see for the HPC / AI community in the next 5-10 years, and how do you see your own skills evolving during this time frame?
We are at a pivot point where the progress of AI-based technologies outpaces predictions. I believe we will be closer to scalable AI-based solutions that are integrated into science workflows and enabling new scientific discoveries at a pace faster than ever before. The skill levels we expect of folks keeps growing as several domains get interconnected. I don’t see that slowing down and expect to see requirements for deep knowledge in several fields to be productive. Personally, I am very excited to live in times where technology growth has the potential to impact science (arguably) faster than ever.
What is your view on the convergence and co-dependence of HPC and AI?
I believe the fields of HPC and AI strongly connected in many ways. Modern AI training and inference are the largest HPC workloads we have seen so far. It is a giant step from a traditional laboratory-scale HPC designed for scientists to global scale HPC for AI designed for every human. We are just scratching the surface on the possibilities here. On the other side, AI training is all about parallel linear algebra – something that the HPC community has been laser-focused on for decades. The convergence of HPC and AI is the perfect match of two fields even better than the HPC and science combination that worked for years.
Do you believe science drives technology or technology drives science?
I believe in a co-evolution between science and technology. New technology has helped improvements in science for decades and some of these effects are quite direct. GPUs are the most recent example where exascale computing using GPUs has enabled accelerating scientific discoveries manyfold. The impact of science on technology is not as obvious in many instances. The simplest somewhat abstract example one can talk about is the impact of modern science on life expectancy and how it has enabled several generations of scientists and technologists to learn from each other directly. There are several other direct examples from designing new materials for batteries, semiconductors as well.
Would you like to share anything about your personal life?
I am a first-generation immigrant to US and generally an optimist on many things including the possibilities in this country, opportunities for science and new technologies. I am your guy if you want to discuss about cricket, chess, Indian food, and why this is the best time and place to live and do fundamental research.
Find your next career at Sandia Labs: sandia.gov/careers
