Visionary retirement threatens particle physics

From our earliest childhood, we have all learned the basics of numbers through a few basic steps. For example, even the greatest minds in the history of mathematics began by counting on their fingers before moving on to addition and multiplication.

As we go further, we begin to encounter somewhat more abstract concepts, such as the equations formulated centuries ago by the Persian mathematician Muhammad ibn Musa al-Khwarizmi. All these tools allowed people to learn more and more advanced mathematical concepts.

These developments themselves led to the emergence of revolutionary new tools, starting with computers. Computers are able to absorb very complex mathematical problems and then find a reliable answer in less time than it takes a human to read an expression.

Inevitably, this technology completely changed all jobs involving complex calculations. In the 1950s, for example, NASA relied on a veritable army of “human computers” that calculated all rocket trajectories using, among other things, a rudimentary calculator. This profession has become completely obsolete with the rise of the computer.

Therefore, we can instinctively say that the problem of the complexity of the equations has more or less disappeared. After all, when there’s a problem, you just need to power up the machine, right? It’s actually more complicated than that. Because the more science advances, the more exponentially the needs in terms of computing power grow.

Particle physics, torture for machines

This is partly the fault of physicists, especially those working at the smallest scales. Indeed, particle physics contains some of the longest equations in existence. And above all, as soon as new parameters are integrated, their complexity increases at a crazy speed. Note that in this context, we can no longer talk about exponential growth; experts talk more about factorial growth.

Suffice it to say, the hope of taming these mathematical monsters with pen and paper is just an illusion… and for a long time, even the most powerful processors have had all the trouble in the world to achieve it.

Indeed, early computers had ridiculous amounts of RAM compared to modern machines. And it was a big problem—literally. Because in order to process an equation, the computer must first store it in its RAM; however, the particle physics equations were too long to be stored entirely in RAM!

As a result, they had to be partially stored on a standard hard disk drive (HDD). Therefore, data had to go back and forth between the hard drive and RAM. This has the effect of further slowing down the solution of these already dauntingly difficult equations.

FORM, a revolution for research…

In order not to find yourself in a technological impasse while waiting for access to more RAM, it was necessary to prefer another angle of attack: software-side optimization. So in the 1980s, a Dutch physicist named Jos Vermassen created a program that would fundamentally change the daily life of researchers.

The program in question, called FORM, makes it possible to bypass this memory problem. Physicist Matt von Hippel explains that this allows us to split these equations into different terms, then assign each of them to a predefined area of ​​the hard disk.

In other words, it allows the processor to treat the HDD as RAM. Thanks to this reorganization controlled by FORM, these huge equations can be treated more quickly.

This technique worked so well that FORM immediately established itself as an indispensable tool in particle physics. And since then, its importance has not diminished. Even today, several scientific papers based on FORM are still published every week. “ Most of the high-precision results obtained by our group over the past 20 years were based mainly on the FORM code. “, explains Swiss physicist Thomas Gehrmann in an interview with Quanta magazine.

And everything indicates that this algorithm will continue to play a decisive role in the future. ” No matter how much memory grows, this concept will remain relevant because there will always be a physics problem that will be limited by the amount of available memory. “, explains Vermaseren’s former student Ben Ruijl, who interviewed Quanta.

… threatens to fall into ruin

However, this situation can now be questioned. Because although they are not physical instruments, logicians are not eternal. As IT changes, so does regular maintenance and updates to keep up with the rest of the ecosystem.

And here physicists face a very, very big problem.

It is indeed an extremely complex and above all highly specialized program; understanding the details of how it works requires extremely advanced computer skills beyond the reach of most particle physicists.

This means that there are not many people who have enough skills in these two areas to maintain FORM. In fact, there are onlyonly one person can do it right now…and this is Vermaseren, the father of software.

However, at the age of 73, the person concerned begins to move towards a decent retirement. But he’s still slow to fully back off, and for good reason: no successor on the horizon yet!

If no one can pick up the torch, the software will become less and less usable, and then completely obsolete. And knowing its importance, this is all that particle physics research can do pay the price for many years.

One of the many perverse effects of the publishing race

The big question is how we got here. How is it that the world’s leading scientific discipline, now in full swing, is so dependent on a single software solution mastered by one person? The answer stems from one of the great challenges of modern research, namely the pressure to publish.

In order to get into prestigious scientific journals, find funding and aim for prizes like the Nobel, you must first gain fame. And of course, that means showing off. For a young researcher, this means publishing a large number of scientific papers that will necessarily be cited by other researchers.

Therefore, certain topics generate significant visibility. On the contrary, other works are almost ignored. After a fundamental measurement, announce the discovery of a new particle and get guaranteed scientific information! On the other hand, exposure to all researchers involved in developing the necessary tools will be much, much less significant.

The conclusion for the young researcher is clear: the king’s way today is to present outstanding experimental results. On the contrary, majoring in instrument maintenance is the best way to end your career in academic oblivion, which amounts to professional suicide for a very ambitious startup.

A system problem that deserves more attention

So it’s easy to see why Vermaseren had such a hard time finding a successor to keep FORM. Admittedly, Ben Ruijl works ” sporadically along with several other former padawans from the creator in the program. But to advance their respective careers, they must also publish their work. It is not possible to focus only on FORM. Therefore, most of the work is still done by pseudo-retirees.

Running out of solutions and probably running out, he took the bull by the horns. Next April, he will organize Extensive colloquium with FORM users to discuss the future. With a bit of luck, this may allow a successor to emerge. Or at least to find ways to save what Matt von Hippel described as ” is one of the most powerful tools of physics “.

But even if a new kid takes the wheel tomorrow, it will only treat the symptoms of a deeper systemic disease. The real problem is the current system that encourages comprehensive publishing at the expense of less important work can be published but equally important in practice.

The FORM case is a particularly vivid example. But there are many other such programs and tools. And the scientific community sometimes tends to take them for granted – wrongly. If this dynamic continues, eventually other disciplines may find themselves in situations comparable to what particle physics is currently in. Therefore, it will be interesting to see whether the institutions leading scientific research on a global scale will accept them alarming warning signs.