recently i saw A fellow particle physicist talks about the calculations that pushed precision to new heights. his tools? A 1980s computer program called FORM.
Particle physicists use some of the longest equations in all of science. For example, to look for signatures of new particles in collisions in the Large Hadron Collider, we draw thousands of diagrams, called Feynman diagrams, representing possible collision outcomes. Each diagram encodes a complex formula that can be millions of terms long. It’s impossible to add up such a formula with pen and paper. Even adding them to your computer is difficult. The algebra rules we learn in school are fast enough for homework, but terribly inefficient for particle physics.
Programs called computer algebra systems attempt to handle these tasks. And if you want to solve the world’s largest equations, one program has stood out for 33 years: FORM.
Developed by Dutch particle physicist Jos Vermaseren, FORM is an essential part of the infrastructure of particle physics and required for the most difficult computations. But like a surprising amount of his critical digital infrastructure, FORM’s maintenance rests largely on his one person, Vermaseren himself. And he’s 73, and he’s starting to move away from developing his FORM. No successor has emerged due to the incentive structure of academia, which emphasizes published papers rather than software tools. If things don’t change, particle physics could slow down dramatically.
FORM began in the mid-1980s, when the role of computers was changing rapidly. Its predecessor, a program called Schoonschip, written by Martinus Veltman, was released as a special chip that plugged into the side of the Atari computer. Vermaseren wanted to create a more accessible program that universities around the world could download. He started programming it in the computer language FORTRAN, which stands for Formula Translation. The name FORM was that riff. (He later switched to his C programming language.) Vermaseren released his software in 1989. By the early 1990s, over 200 of his institutions around the world had downloaded it, and the number continued to grow.
Since 2000, a particle physics paper citing FORM has been published on average every few days. “largely [high-precision] The results obtained by our group over the last 20 years were largely based on the FORM code,” said Professor Thomas Gehrmann of the University of Zurich.
Some of FORM’s popularity stems from specialized algorithms built over the years, such as tricks for quickly multiplying certain parts of Feynman diagrams, and procedures for rearranging equations to make as few multiplications and additions as possible. increase. But FORM’s oldest and most powerful advantage is how it handles memory.
Just as humans have two types of memory, short-term memory and long-term memory, computers have two types of memory: main memory and external memory. Main memory (her RAM in the computer) is easily accessible on-the-fly, but limited in size. External memory devices such as hard disks and solid state drives can hold much more information, but are slower. Solving long equations should be stored in main memory for easy manipulation.
In the 80’s both types of memory were limited. “FORM was built in a time when there was little memory and no disk space. said Ben Ruijl, a postdoctoral researcher at the Institute. technology Zurich. This posed a challenge: the equation was too long for him to handle in memory. To calculate this, the operating system had to treat the hard disk as if it were main memory. The operating system doesn’t know how big the equation is expected to be, and stores data in a collection of “pages” on the hard disk, switching between pages as often as needed. This is an inefficient process called swapping.
FORM bypasses swapping and uses its own technique. When working with equations in FORM, the program allocates a certain amount of space on your hard disk for each term. This technique allows the software to more easily track where the parts of the equation are. And you can easily put those parts back into main memory when you need them without having to access the rest.
Memory has increased since the early days of FORM, a million-fold improvement from 128KB of RAM in a 1985 Atari 130XE to 128GB of RAM on my powered-up desktop. But the trick Vermaseren developed is still important. As particle physicists examine petabytes of data from the Large Hadron Collider looking for evidence of new particles, the need for precision grows and the length of the equations grows.
