Scientists are using Apple technology in a major global project to help improve man’s understanding of the structure of the universe. Intel-based Macs will play an important role in the ALICE experiment at the CERN international particle physics laboratory in Geneva, helping to measure what happens when particles collide with unprecedented force at close to the speed of light.

The ALICE experiment is one of five detector experiments that will exploit collisions produced by CERN’s Large Hadron Collider (LHC), the world’s largest particle accelerator, which will probe deeper into matter than ever before when it becomes operational in 2007. The results from ALICE and the other detector experiments will help the LHC test fundamental scientific assumptions about matter and anti-matter, and is likely to permanently alter our views about the universe.

To achieve their results, the five experiments will handle as much information as the entire European telecommunications network. They will produce close to 10 petabytes of data a year — equivalent to a stack of CDs around 20 km high.

“Now that Apple has moved to Intel processors, we see real opportunities to use the Mac as our main computational engine”, says ALICE Computing Coordinator, Federico Carminati. “Apple is a very nice solution which offers two environments in one system — UNIX processing power and the world of office applications”.

Computing centres worldwide

ALICE — which is an acronym for A Large Ion Collider Experiment — is tasked to detect the collision of heavy elements in the LHC, particularly lead and hydrogen nuclei (protons). Its construction challenges are immense. The LHC uses superconductivity to guide particles along a curved path, which means its magnets have to be cooled to -271 degrees, or close to absolute zero.

The ALICE team of physicists will have vast amounts of data to analyse throughout the project’s 10-year lifespan, and will depend heavily on computer technology. “All 1000 physicists on the ALICE team need to use computers — varying from 10% of their time to the whole of their working day”, explains Carminati. “To analyse the data and reap the physics from it you need powerful computers to do mass statistical analysis”.

Now that Apple has moved to Intel processors, we see real opportunities to use the Mac as our main computational engine.

— Federico Carminati, ALICE Computing Coordinator

The ALICE computing framework has taken years to devise, because of the need to accommodate the simulation of heavy-ion and proton-proton interactions, the resulting detectors’ response, and the reconstruction and analysis of the data coming from simulated and real interactions.

For their specialist research requirements, the physicists need to be able to develop their own software to meet their specific analysis needs. A software infrastructure is currently being built to ensure their bespoke development activities use consistent and cost-effective methodologies.

Data from the ALICE experiment will be stored in a heterogeneous network of computing centres worldwide, funded by national partner organisations. The ALICE offline team is managing access to this data and interaction with about 40 centres around the world.

Sharing knowledge

Federico Carminati’s computing group at ALICE have two main objectives in selecting the development and processing platforms to be used at CERN and across the computing centre grid network. First, because of the volumes of data to be processed, they are seeking to minimise costs by implementing solutions offering best price / performance ratios.

“By 2009, we think we will need 30M SPECints of computing power”, says Federico Carminati. “Buying computing is the responsibility of each computing centre, including CERN computing centre. However, we do participate in the specification of the platforms we want our software to run on. That means we have a big responsibility to ensure the platforms deliver the best deal for all our funding partners”.