Francesco Massimo: advancing laser-plasma acceleration simulation with Smilei open-source code

“The Smilei code was created in 2013. The initial development team involved researchers from the laboratory for the use of intense lasers (LULI - CNRS/École Polytechnique/Sorbonne Univ.), the Leprince Ringuet Laboratory (LRR - CNRS/École Polytechnique) and the House of Simulation (MdS - Univ. Paris-Saclay/UVSQ/Inria/CNRS/CEA). As it is the case today, at that time there was a high demand of numerical modeling of laser-plasma interaction,” comments Francesco Massimo, now a specialist in this code at the Gas and Plasma Physics Laboratory (LPGP - Université Paris-Saclay/CNRS). “The team, mainly composed of very young researchers, quickly realized the importance of having an open-source multi-purpose simulation code. This explains why now Smilei is used in several research fields, like particle acceleration, astrophysics or quantum electrodynamics,” adds the researcher who has started working on this code since 2017 and his arrival at the LRR.

The power of open source

“Imagine if people could use or modify this code to advance knowledge. How fascinating would that be? This is the power of open-source codes. However, there are myriads of challenges to overcome,” says Francesco Massimo when asked about his decision to opt for open source. “The code is not funded through the licenses that users pay for. Yet the amount of resources needed is significant and often underestimated, both in terms of finances and skills required to develop, verify and maintain the code. Fortunately, Smilei is the result of a collaboration between experts in high-performance computing and plasma physics from several laboratories on the Saclay plateau. This is one of the aspects that has allowed the code to thrive. And very recently, I have started to see researchers including the input file of their calculation in the article itself, in order to address the question of reproducibility.” In June 2024, there were at least 190 peer-reviewed articles describing simulations performed with the help of Smilei. “There are even researchers working in fields that we were unaware of who are using the code,” enthuses the LPGP researcher.

The quest of research, from Italy to France

“I’ve always been fascinated by physics,” recalls Francesco Massimo when explaining his career path. “When I was doing a bachelor’s degree in electronic engineering at Sapienza University in Rome, which included a focus on applied electromagnetism, my friends who were studying physics were talking to me about their courses on the numerical resolution of differential equations, which aroused my curiosity. And thanks to my thesis supervisors, I got hooked on the numerical solution of differential equations of mathematical models in physics. During my PhD, I had a first-hand experience of collaborative code development for particle acceleration in plasmas.

After his thesis, the young researcher undertook several post-doctorates in plasma and particle physics laboratories on the Saclay plateau. There, he focused mainly on implementing various physics modules within the Smilei code and improving the scalability of its performance. “Starting to contribute to Smilei was a natural evolution of my previous activities,” says the researcher who joined the LPGP in 2022 when recruited by CNRS. With this code, he tries to describe in a self-consistently way the interaction between intense laser pulses and plasmas, under experimental conditions of interest to his field. “Plasma physics in these regimes is highly non-linear. This means that the slightest change in a single parameter can sometimes completely change the physics. This makes it difficult not only to simulate the phenomena under more realistic conditions, but also to understand its complexity,” explains Francesco Massimo when asked about the particularity of his research field.
 

Impactful research and open science

“While publishing their research results through scientific articles remains a major activity for researchers, it should be noted that the number of publications is not the only indicator of a high-quality scientific work. Other criteria of paramount importance include the reproducibility of these results, and the availability of tools and data that can be re-used by the scientific community,” Francesco Massimo points out. “This is something that the research institutions, funding agencies and evaluators should take into account when assessing the career of a scientist.” For him, researchers are not just publication machines but intermediaries between science and society. As such, Francesco Massimo favours the principle of high-impact (immediate or future) and reproducible research over the number of publications. “Publishing numerous articles/codes without documentation or without all the information needed to reproduce the results doesn’t help the scientific community very much. It is therefore extremely important to train PhD students, as well as permanent researchers on the best open science practices.”
 

Planting and watering the seeds of open science for a stimulating future

When it comes to the future of scientific research in terms of open source, the researcher is optimistic. “For students, having access to an open-source simulation code provides them with the opportunity to explore in a more interactive manner the phenomena they have been learning about since childhood in physics lessons. Images and animations from numerical simulations are often effective means to arousing their interest and encouraging them to study the underlying physical models.”

As for the Smilei code, the researcher is teeming with ideas for physics modules to integrate. “Other modules and geometries that are already functional on a central processing unit (CPU) will soon be functioning also on a graphics processing unit (GPU). One of our objectives is it to obtain good energy performance, and this porting work will enable simulations to be carried out on the most recent and efficient architectures, which are now often based on GPUs.” Looking for future collaborations is another of Francesco Massimo's concerns. “We have already established a strong user base in several countries, whose feedback has greatly contributed to improving the code. I would be delighted to play an active role in its expansion into new horizons,” he concludes.