Workload Requirement

  • An HPC solution capable of powering Kalman Filter workflows for particle physics.

    • Technical Result

  • Machine-Learning Solution capable of meeting the fast-increasing compute needs of particle physics.

    • Solution At A Glance

  • Intelligence Processor Unit (IPU) server appliance solution.

    • Business Result

  • Ability to process faster, more detailed particle physics computational workloads.

    • Purpose built processors are being increasingly used for parallel processing workloads, which involve processing huge amounts of data. Within scientific research, particle physics is one of these domaines.

    A generative adversarial network (GAN) is a class of machine learning framework designed where two neural networks contest with each other in a game (in the form of a zero-sum game, where one agent's gain is another agent's loss).

    Given a training set, this technique learns to generate new data with the same statistics as the training set. Though originally proposed as a form of generative model for unsupervised learning, GANs have also proven useful for semi-supervised learning, fully supervised learning, and reinforcement learning. GANs have been trained to accurately approximate bottlenecks in computationally expensive simulations of particle physics experiments. Applications in the context of present and proposed CERN experiments have demonstrated the potential of these methods for accelerating simulation and/or improving simulation fidelity.

    The IPU allowed for models to be trained faster than on CPU and GPU appliances.

    Popular destinations for deployment include the Nordics for low-cost, 100% sustainable power.

    We can arrange a remote testing facility and arrange a proof of concept loan unit for select customers - feel free to get in touch by telephone on +(0)207 352 7007 or by email at enquiries@bsi.uk.com

    A white paper from Graphcore and the Physics Laboratory at the University of Bristol is available to download here -