CGI is building a business around connecting quantum processors to existing enterprise systems. Its work spans optimization, security and federal logistics while the underlying hardware remains outside the company.
CGI is not trying to build the next quantum processor. It is trying to make quantum hardware usable inside organizations that already depend on legacy software, supply chains and government systems. The strategy became more explicit in September 2026, when CGI and D-Wave announced a strategic partnership under which CGI would include D-Wave Advantage2 and hybrid solvers in client offerings for logistics, transportation and retail.
That distinction defines the strategy described by Victor Foulk, Vice President of Emerging Technology at CGI Federal, and Curtis Nybo, Global Quantum Computing Lead at CGI. CGI has roughly 90,000 consultants delivering end-to-end IT services across commercial enterprises and federal agencies. Its quantum role sits at the point where experimental hardware meets operational software: identify a difficult business problem, select an appropriate quantum architecture and connect the result to systems that employees already use.
The partnership is therefore an enterprise-integration agreement, not a processor-development program. CGI is positioning Advantage2 and D-Wave hybrid solvers inside an existing service portfolio rather than asking clients to purchase, house or operate quantum hardware themselves. That model lowers the organizational barrier to experimentation, much as cloud computing allowed companies to rent specialized infrastructure without building a data center.
The approach avoids asking a client to become a quantum hardware specialist. Through partnerships and subscriptions to AWS and Azure, CGI teams can access providers using both quantum annealing and gate-based architectures. The intended abstraction is practical rather than mystical. A user should not need to know whether a calculation ran on a classical CPU, a GPU or a quantum processing unit; the relevant test is whether the combined system produces a better answer or reaches one faster.
That is a systems-integration proposition rather than evidence of quantum advantage. The available material provides no processor benchmark, no qubit count and no independent comparison showing that a quantum workflow has outperformed a classical alternative. In scientific terms, a deployment announcement is not a performance study: a credible advantage claim would require a defined workload, a strong classical baseline, reproducible measurement conditions and uncertainty estimates.
In optimization, many enterprise problems can be expressed as quadratic unconstrained binary optimization or related Ising formulations. A quantum annealer searches an energy landscape representing candidate solutions, while classical software commonly performs constraint handling, preprocessing, parameter selection and post-processing. Hybrid solvers divide the workload between those components, so the useful unit of analysis is the complete algorithmic pipeline rather than the quantum processor in isolation.
CGI says client interest has moved from broad exploration toward applied use cases. Optimization is the clearest entry point because logistics, manufacturing and healthcare routinely involve routing, scheduling, allocation and other problems that change as conditions change. The D-Wave partnership gives CGI access to the company's optimization tools and annealers, which CGI says can operate with 99% uptime for continuous background re-optimization.
That figure describes the reported availability of systems used through the partnership. It does not establish that every optimization is faster, cheaper or more accurate than a classical method. Nor does it turn quantum annealing into a universal gate-based quantum computer. The practical question is narrower: whether a selected workload benefits from repeatedly updating a solution within a wider classical workflow.
The distinction matters because quantum annealing and gate-based quantum computing use different physical and computational models. Annealing is designed around optimization landscapes and adiabatic or energy-minimization principles, while gate-based machines manipulate quantum states through programmed operations. Research communities at MIT and elsewhere continue to investigate when either model can offer a measurable computational benefit, but no general rule says that a quantum device improves every difficult optimization problem.
CGI is also exploring computational chemistry with manufacturers of physical materials such as paint. The proposed use is to model how a material degrades over time under particular environmental conditions, potentially reducing reliance on lengthy laboratory testing. The material supplied here does not report a completed quantum simulation or a measured improvement in predictive accuracy, so this remains an application area under exploration rather than a demonstrated industrial result.
Materials simulation is scientifically plausible because molecular energy estimation is one of the areas in which quantum algorithms may eventually complement classical methods. Yet practical chemistry workflows also require accurate physical models, sufficient circuit depth, error mitigation or correction, and validation against laboratory measurements. A proposed simulation is not equivalent to a validated product prediction, and no such validation is reported here.
In the public sector, CGI Federal has established a fiscal year 2027 research and development agreement with the US Defense Logistics Agency, the largest military logistics agency in the government. Working with strategic industry partners and an onshore delivery center in Knoxville, Tennessee, the company is developing applications for warehouse management, reverse supply-chain disposition and contested logistics.
Related materials describe a joint Quantum Pathfinder initiative involving CGI Federal, D-Wave and the Defense Logistics Agency. Its focus is inventory management, allocation of supplies and reverse logistics, with the first detailed application scenarios expected to be refined during the following months. This is an applied test of how quantum and classical tools might enter a logistics mission, not evidence that a quantum processor has already transformed military distribution.
The reported target is operational improvement rather than a new quantum machine. CGI says that relatively small changes to delivery routes or supply-chain logic can produce a 10% reduction in delivery costs and lower maintenance overhead. The material does not identify a completed deployment, a measured quantum contribution or a controlled comparison establishing that the stated reduction came from quantum computing. Those limits matter: a projected or application-level benefit cannot be presented as a verified hardware result.
Production integration creates its own technical workload. A quantum application must receive parameters securely from existing data systems and return results through dashboards, reports or decision engines. An annealing system might support continuous optimization in the background, while a gate-based system could appear in practice as a research team running targeted simulations within a larger classical process. Data-transfer latency, API reliability, queueing, calibration and model maintenance can all affect the final business result.
Victor Foulk has framed this transition as a need to build a coalition among government, industry and academia and to buy outcomes rather than machines. That language reflects a procurement principle familiar from other advanced technologies: an agency should acquire a measurable mission capability, not merely access to an impressive technical demonstration. The same logic is visible in NASA and CERN programs, where instruments are judged by the scientific or operational results they enable within a much larger system.
For many organizations the first quantum-related project is defensive. CGI's cybersecurity teams are helping clients locate cryptography embedded across large enterprise data estates and migrate legacy systems toward NIST-approved post-quantum cryptography algorithms. This work concerns conventional computers preparing for future cryptographic risks; it is not the same as using a quantum computer to protect a network. Organizations can begin with NIST's PQC guidance while cataloguing algorithms, certificates, dependencies and replacement timelines.
Foulk has described a practical federal sequence: first inventory the cryptography, then assess the lifetime and sensitivity of the data it protects. That prioritization is important because post-quantum migration is not a single software upgrade. Long-lived secrets, archived information and systems with difficult replacement cycles may require earlier action than data with a short operational life.
Security migration can expose a second question. Once an organization has mapped where cryptography sits and considered why quantum computing matters to encryption, it may begin examining optimization or simulation opportunities. The link between the two activities is organizational discovery, not proof that current quantum systems can break deployed RSA encryption.
Federal adoption adds sovereignty and procurement constraints. Secure hybrid classical-quantum infrastructure may need to be housed domestically, managed by US personnel and connected through encrypted links and managed certificates. Acquisition staff do not need to design qubits or operate cryogenic systems, but they do need to understand which workloads a system handles well and how that system fits securely into an existing mission architecture.
Quantum security planning also requires separating the threat model from the hardware roadmap. Fault-tolerant gate-based machines capable of running long algorithms would require extensive error correction because physical qubits are vulnerable to noise and imperfect operations. The scale and timing of that engineering challenge remain uncertain, which is why migration planning should be based on data sensitivity and system lifetimes rather than on a single forecast of when a particular machine will arrive.
CGI's own framing is therefore more credible when it is measured in integration and mission outcomes than in qubit counts. Its federal practice uses the term Mission Return on Investment to describe lower execution costs and faster delivery rather than ordinary commercial profit and loss. The same practical emphasis appears in an earlier quantum strategy report on how a financial institution is combining security migration with hardware exposure before quantum finance has arrived.
Clients also face change-management costs that hardware demonstrations rarely show. An organization may need to retire classical optimization models built through years of internal investment, train stakeholders during application development and verify that quantum outputs do not disrupt established decision processes. CGI's use of agentic digital-workforce platforms adds another layer: AI can inspect legacy IT, separate a large task into components and route only selected subproblems to quantum hardware while classical systems handle the rest.
The evidence supports a clear reading of CGI's position. It is acting as a broker and systems integrator between quantum providers and organizations with difficult workflows, not as a manufacturer of qubits and not as a demonstrated source of general quantum advantage. That may be the right commercial role at this stage, because the immediate obstacle is often not access to a processor but proving that a processor can improve a complete workflow after data movement, classical processing, security, training and maintenance are counted.
A hybrid quantum-classical system combines different computing resources rather than replacing one with another. The quantum component may handle a narrowly defined optimization or simulation subtask while classical machines prepare inputs, manage control and interpret outputs. For CGI's model to produce durable value, each claimed improvement must therefore be measured at the workflow level against a strong classical baseline, with transparent metrics such as solution quality, runtime, energy use, total cost and failure rate.
Until those measurements are made public, the company's quantum activity is best understood as disciplined preparation for commercialization rather than commercialization already achieved. CGI's partnership with D-Wave shows how quantum capability may reach enterprises through consulting, cloud access, hybrid algorithms and mission-specific integration. It does not by itself establish quantum advantage, but it identifies a realistic route by which future improvements could be tested in logistics, transportation, retail, materials and federal operations.
