WIXLIB

QuantumLearning Machine

Get Ready for the Quantum RevolutionTo tackle increasingly demanding industry challenges, we need to findalternatives to existing computing solutions. In this context, QuantumComputing is one of the most promising technologies.Leveraging quantum propertiesAtos Quantum R&D ProgramThe first quantum revolution has took place at a microscopic leveland brought some major innovations of the 20th century such asthe transistor, the laser, superconducting devices and optical fibers.Many experts are convinced that the second quantum revolution,which will see quantum information emerging, is already upon us.A binary digit is always in one of two definite states (0 or 1) howeverquantum computing uses qubits which can be in a superposition ofstates (1 and 0 superposed), enabling massively parallel computation.This multiplies capacity, so that quantum computers with a modestnumber of qubits, as little as 60, could solve problems that today’smost powerful supercomputers cannot address with years ofcomputing time! Advances in quantum science are extensive anddemonstrate the reality of the tremendous opportunities to acceleratecalculations, which together allow quantum superposition andquantum entanglement.Atos Quantum’ is the first quantum computing industry program inEurope. This global program aims to develop quantum computingsolutions but also to learn how to enhance cyber security products toanticipate quantum advantage and its impact on cryptography.Atos’ ambition is to be a quantum player in three domains: quantumprogramming and simulation platforms and, later, next-generationquantum-powered supercomputers, as well as quantum-safecybersecurity. It is piloted by a Scientific Council with world-renownedmembers. The ‘Atos Quantum’ Scientific Council is enriched with theimmense knowledge and visionary power of its members, all highlyrenowned researchers. The Council is chaired by the Chairman andCEO of Atos and is made up of universally recognized quantumphysicists and mathematicians.The first products of the Atos Quantum Program have been the AtosQuantum Learning Machine - Atos’ world-class quantum simulationappliance - and the Atos Quantum consulting practice.Go beyond HPCQuantum Computing is a new computing paradigm that will solve critical problems in all industries more efficiently than current highperformance computing (HPC) systems.Public Sector &DefenceManufacturing Autonomus vehicleLogiasticSupply chainSoftware validationBatteriesPolymer Neural networksProcess optimizationCryptanalysisMaterial scienceNanotechnologiesTelecom, Media &TechnologyFinancial Services& Insurance Fraud detectionTrading strategiesMarket simulationPortfolio optimizationRisk assessmentCryptocurrencyResources &ServicesHealthcare & LifeSciences Personalized content Smart grids Genomics 5G antenna location Flight scheduling Virtual Screening Chip layout optimization Oil well optimization Protein folding Post-quantum cryptography Yeild management Drug discovery Cybersecurity Personalized medicine Carbon dioxide captureAtos QLM benefits Bootstrap in Quantum Computing Ease end-user quantum programminglanguage, a set of provided QuantumLibraries and a Jupyter Notebook environment Program and execute hybrid quantumclassical algorithms Support real noise simulation models ondifferent hardware Integrate with existing frameworks toleverage algorithms developed via otherframeworks Simulate different technologies throughits hardware agnostic environment withunprecedented performances. Explore both digital and analog quantumcomputing technological approaches02

Develop quantum applications todayThe Atos Quantum Learning Machine (Atos QLM) is a complete onpremise environment designed for quantum software developers.It is dedicated to the development of quantum software, trainingand he Atos Quantum Learning Machine provides 3 environmentsIt embeds a programming platform anda high-performance quantum simulator.Its capabilities and performances areunmatched on the market. The AtosQuantum Learning Machine allowsresearchers, engineers and students todevelop and experiment with quantumsoftware. Powered by a powerful dedicatedhardware infrastructure, the Atos QLM willemulate execution as a genuine quantumcomputer would.Software developed on the Atos QLMcan run on simulated, as well as on futurequantum accelerators, without changing aline. The Atos Quantum Learning Machinecomputes the exact execution of a quantumprogram, with double digit precision. Itsimulates the laws of physics, which are atthe very heart of quantum computing.This is very different from existing quantumprocessors, which suffer from quantum noise,quantum decoherence, and manufacturingbiases, as well as performance bottlenecks.Simulation on the Atos Quantum LearningMachine enables developers to focus on theirapplications and algorithms, without havingto wait for quantum machines to be available.The Atos Quantum Learning Machine is a complete applianceThanks to the bespoke software andhardware developed by Atos, the Atos QLMhas superior simulation capabilities, muchmore than any other affordable appliance. Inits maximum configuration, it can simulateup to 41 Qbits, which was until now, onlypossible on large supercomputers, despitethe fact that the Atos QLM has the physicaldimensions of a simple business server.The appliance is composed, nonexhaustively of: A high-level quantum hybrid language, builton top of the popular Python language A specific hardware infrastructure, withlarge in-memory capacity and – availablesoon – a dedicated hardware accelerator A set of mathematic libraries andalgorithms – ready to use An extensible quantum circuit model (datarepresentation model) A universal quantum assemblyprogramming language (AQASM, AtosQuantum Assembly Language)Thanks to Atos’ large in-memoryinfrastructure, it is simple and seamless toupgrade the simulation capabilities from 30to 41 qubits. Grow as your business needs itwhile leveraging existing investments.The Atos Quantum Assembler (AQASM)The Atos Quantum Assembler is an important component of the Atos QLM appliance. AQASM is designed in the Atos R&D labs. AQASM isAtos’ implementation of the universal quantum software development language: QASM. Programs using AQASM can be used on the quantumsimulator today, as well as on the physical quantum computers of tomorrow. AQASM is configurable; it has the ability to program new quantumgates or to mix existing gates. AQASM offers high-level wrappers for developers and accepts quantum programs from other frameworks thanksto a versatile software development π/4)PH(π/8)HPH(π/2)PH(π/4)HPH(π/2)HQuantum Learning Machine03

Atos Quantum Learning Machinefunctional scopeProgrammingQPUOptimizationAQASMpyAQASMQPUGate set rewriterPluginsAssembly language tobuild quantum circuitsPython extension toAQASMQuantum processingunit emulationAbstract gatesExpressive andconcise programmingCIRCQLIBCircuit optimizersBinary format ofquantum circuitsQuantum arithmeticlibrariesGeneric circuitoptimizerTopology constraintsolverSimulationQuantum algorithmsVariational Algorithms,,such as QAOA andVQESimulatorsPhysicsDifferent simulationmodules for CircuitmodelsPhysical noise modelsGateset tomographySimulated QuantumAnnealingQuantum Noise Simulation0.5Ideal Quantum CPUNoisy Quantum CPU0.4ProbabilityThe simulation of physics includesdifferent models for quantum noise.Should the underlying technology betrapped ions, superconducting circuitsor semiconducting silicon, the noisemodels can be chosen and used to be asclose as possible to reality.0.30.20.10.0051015202530kA broad range of services based on the AtosQLMJoin the Quantum revolution with Atos expertsAtos provides a broad range of services based on theAtos QLM. Our mission, with our Quantum experts, is toaccompany you in every aspects.From the discovery of the Atos QLM appliance, to thedevelopment and simulation of your first Quantumcodes, our experts’ outstanding teaching skills and theQLM intuitive programming environment gives you theknowledges to join the Quantum revolution04

Atos QLM E – accelerating research in theNISQ eraNISQ – Noisy Intermediate-Scale Quantum – devices will be the first quantum accelerators to becommercialized in the next few years. Herein lies a double challenge for the industry: developingNISQ algorithms is as important as building the machines, since both are required to identifyconcrete applications. Atos QLM E has been optimized to drastically reduce the compilation timeof hybrid quantum-classical algorithms simulations, leading to quicker advancement in applicationresearch.Get your quantumsimulations results up to 12xfasterOptimized for variationalalgorithms best suited for theNISQ era (up to 30 qubits)Upgrade your QLM to a QLME very easily to benefit fromthese enhanced featuresAccelerate research on variational algorithms using Atos QLM EQuantum ApproximateOptimization AlgorithmVariational QuantumEigensolverVariational QuantumFactoringVariational QuantumClassifier among other application fields !Explore both analog and digital quantumsimulation paths to solve optimization usecases using the Atos QLMNP-hard optimization problemshave critical applications in allindustries, such as portfoliomanagement, logistics, antennalocation, chip design or clinicaltrial database search. Classicaltools like solvers and heuristicsprovide either exact solutionsfor a small number of variablesor approximate ones for largerproblems, but their use isproblem-dependent, and they donot all have provable behaviors.In this field, quantum computingholds major promises: Address larger problems Improve accuracy level Reduce time to solution And thus, reduce energyconsumptionUsing the Atos QLM, you canexplore now two quantumrelated technological pathsto solve such problems in thefuture:Key benefits of using both approaches in theAtos QLM Better understand the specificities of digital and analog quantumcomputing Compare the benefits and constraints of two quantum computingtechnological paths applied to combinatorial optimization problems Prepare code to then run on NISQ accelerators or QuantumAnnealing machines using the same programming environmentDescribe your use case as ahigh-level NP-hard problemEncode it in a mathematicalformulation Max cut QUBO (QuadraticUnconstrained BinaryOprimization) formulation Graph partitioning Graph colouring K-Clique Vertex Cover Number Partitioning Binary Integer LinearProgramming Ising modelSimulate it Digital Quantum Computing:Variational Algorithmssuch as QAOA (QuantumApproximate OptimizationAlgorithm) Analog QuantumComputing: SQA (SimulatedQuantum Annealing) orSBA (Simulated BifurcationAlgorithm)Quantum Learning Machine05