An IT RD Collaboration Framework Quantum Computing for

An IT R&D Collaboration Framework Quantum Computing for HEP Workshop Alberto Di Meglio – CERN openlab Head 05/11/2018 1

DRIVING INNOVATION SINCE 2001 CERN openlab Collaboration Board 2017 CERN openlab @ QC 4 HEP 2

CERN OPENLAB’S MISSION Our recipe for success Evaluate and test state-of -the-art technologies in a challenging environment and improve them in collaboration with industry. Collaborate and exchange ideas with other communities to create knowledge and innovation. Train the next generation of engineers/researchers, promote education and cultural exchanges. Communicate results, demostrate impact, and reach new audiences. CERN openlab @ QC 4 HEP 3

Collaboration Model Pick a challenging problem Requirements Check how this problem impacts or is impacted by Industry and Research Assessment Due Diligence If a good match is found, define a joint project with a company and/or academic institute Technical Negotiation Company invests because can get a return from the investment Legal/Financial Negotiation

JOINT R&D PROJECTS High-bandwidth fabrics, accelerated platforms for data acquisition Fast simulation, Data quality monitoring, anomaly detection, physics data reduction, benchmarking/scalability, systems biology and large-scale multidisciplinary platforms Predictive/proactive maintenance and operations Data Analytics, Machine Learning (many) Data Acquisition (LHCb, CMS, Dune, IT-CF) Code modernizati on (EP-SFT, IT-CF, OPL) Control Systems (BE-ICS) Software Defined Networks, Security Cloud infra (IT-CM) Networks (IT -CS) Data Storage (IT-ST, IT-DB, EP-DT) CERN openlab @ QC 4 HEP Simulation, HPC on the Cloud, benchmarking Cloud federations, containers, scalability Storage architectures, scalability, monitoring 5

ICT WHITE PAPER 2017 Published on 21 September 2017 Summarises the results of extensive consultations with CERN experts, LHC Experiments representatives, ICT companies and international research labs and academic institutes Based on 4 major research topics and 16 challenge areas A solid starting point for a very challenging and constructive Phase VI

Three Main Areas of R&D Increase data centre performance with hardware accelerators (FPGAs, GPUs, . . ) optimized software COMPUTING CHALLENGES Scale out capacity with public clouds, HPC, new architectures Change the computing paradigms with new technologies like Machine Learning, Deep Learning, Advanced Data Analytics, Quantum Computing CERN openlab @ QC 4 HEP 7

Artificial Intelligence (ML/DL) Convolutional Neural Networks and Image Analysis Very-Fast-Rate Data Acquisition Processors Generative Adversarial Networks Physics Collisions Track Reconstruction Deconvolutional Neural Network and Networks Interpretation Fast Physics Simulation Predictive Maintenance for the Accelerator Complex Natural Language Processing Environmental Control Systems, Better Mobility, Knowledge Discovery Platforms Chatbots and Smart AI Assistants CERN openlab @ QC 4 HEP 8

Disruptive Architectures: Neuromorphism and Quantum Devices Always look with keen interests at new architectures as a way of disrupting (for the better) how we do things Actively following developments in Neuromorphic Computing and Quantum Computing Which use cases from HEP workflows would benefit? Real-time DAQ filtering, Track reconstruction, radiation simulation, basic particle interactions, QCD. And non-HEP workflows as well: brain simulation CERN openlab @ QC 4 HEP 9

KNOWLEDGE SHARING Working with communities beyond high-energy physics Working closely with CERN KT on initiatives aimed at transferring tools, skills, and knowledge from the high-energy physics community to other research fields. Aligned with the recently approved knowledge transfer strategy for the benefit of medical applications CERN openlab @ QC 4 HEP 10

IP Management, Open * • The basic principle of any CERN openlab collaboration is openness • We assume shared IP of results among project members • More specific IP agreements can be discussed with the CERN KT Office • Within the respect of limited confidentiality agreements and short embargo periods, we expect the results of the projects to be released to the scientific communities following open policies

MEASURING IMPACT Communication and Outreach Our website: http: //openlab. cern/ ~150 k unique visitors/year (+100%) Articles in other CERN and external channels ~900 press cuts (+300%) Facebook: http: //cern. ch/go/p 7 p. F 3. 5 k people reached / 100 k single posts Facebook group: http: //cern. ch/go/n 6 x. D Twitter: http: //cern. ch/go/7 vvk Workplace: http: //cern. ch/go/6 MTQ Linked. In: http: //cern. ch/go/NK 9 k Join the alumni: https: //alumni. cern/ CERN openlab @ QC 4 HEP 12

SUMMER STUDENT PROGRAMME 2000 1840 1800 1540 1600 90 80 1580 1479 70 1400 60 1200 50 1000 850 40 39 750 800 25 600 490 15 400 15 200 15 15 330 22 37 40 40 30 23 20 15 10 92 0 In 2018 § 1840 applicants § 40 selected students § 14 lectures § Visits to external labs and companies § Lightning talks session § 40 Technical reports 0 2006 2007 2008 2009 2010 2011 2012 Candidates 2013 2014 2015 2016 2017 2018 Selected CERN openlab @ QC 4 HEP 13

CONTACTS ALBERTO DI MEGLIO CERN openlab Head alberto. di. meglio@cern. ch ANDREW PURCELL CERN openlab Communications Officer andrew. purcell@cern. ch MARIA GIRONE CERN openlab CTO maria. girone@cern. ch KRISTINA GUNNE CERN openlab Administration/Finance Officer kristina. gunne@cern. ch FONS RADEMAKERS CERN openlab CRO fons. rademakers@cern. ch FEDERICO CARMINATI CERN openlab CÌO federico. carminati@cern. ch www. cern. ch/openlab CERN openlab @ QC 4 HEP 14

Enjoy the Workshop! Thanks! alberto. di. meglio@cern. ch @Alberto. Di. Meglio CERN openlab @ QC 4 HEP 15