In the year 2000, the HEP community decided to adopt the recent concept of computing grid for the computing of the experiments at the Large Hadron Collider at CERN (LHC). It took a decade of intense work to turn the Grid vision to reality. It was also a successful decision, as the worldwide LHC computing grid (WLCG) has processed all LHC data and it has been the computing platform that has allowed the discovery of the Higgs Boson, the so far, the crowning achievement of LHC.
One of the major problems facing the designer of the GRID was the optimal usage of its resources. Attempts at simulating this complex and highly non-linear environment did not yield practically usable results, and the development of a central “optimiser” for job and space management never achieved the expected performance. For job submission and management, a solution was implemented, based on a local optimisation of the workload, which lead to a satisfactory overall utilisation of the resources.
The same cannot be said for data, which is either colocated manually or distributed according to some “reasonable” heuristics, such as topological distance between the data producer and the storage element. While this has given acceptable results, we do not know how far we are from the optimal data distribution, nor how many resources could be saved or how much the performance could improve with a better data placement strategy. This is because the attempts to increase the efficiency of the space management that have been made so far are based on the collection and analysis of the data through monitoring systems, but no action is taken in real time, thus limiting the efficiency of the whole system.
With the foreseen increase in the data produced by the LHC experiments, and the relative tightening of the available resources, for which the growth is constrained by a constant budget, a better data placement strategy becomes extremely relevant for the scientific success of the future LHC scientific programme. The sheer size of the WLCG Grid (42 countries, 170 computing centres, 2 million tasks run every day, 800,000 computer cores, 500 PB on disk and 400 PB on tape) makes such an optimisation mandatory and the expected gain will be relevant.
While this problem has shown to be intractable till now, quantum computing (QC) will substantially improve the optimisation of the storage.
This project will develop quantum algorithms for the optimisation of the storage distribution on the WLCG Grid, starting from the specific case of the ALICE Collaboration at LHC. We will try to determine the optimal storage, movement and access of the data produced by the ALICE experiment in quasi real-time, in order to improve resource allocation and usage and to increase the efficiency of data handling workflow.
The work will be done in collaboration with Google, which has extensive experience in distributed computing and is developing one of the most advanced hardware and software QC programmes.