The financial services industry has always been a heavy consumer of computational power. From calculating derivatives and forecasting risk to detecting fraudulent transactions, the sector thrives on the ability to process massive volumes of data quickly and accurately. Traditionally, this reliance has been served by high-performance computing (HPC) clusters and, more recently, GPU-accelerated systems. However, now another frontier is opening up: quantum computing.
While true large-scale, fault-tolerant quantum systems are still years away, global banks and fintech vendors are not standing still. They are running pilots, partnering with technology providers, and experimenting with both quantum algorithms and supercomputing infrastructures to gain an early edge. As the World Economic Forum notes, projects are already underway in areas such as fraud prevention, portfolio optimization, and advanced risk modeling.
In this blog, we explore the most promising use cases, highlight real-world pilots, and examine the vendors shaping this emerging space.
Why Fintech Needs Quantum and HPC Power
The financial sector is a natural candidate for quantum and supercomputing adoption because:
- Data complexity is exploding. From payments to capital markets, financial firms generate terabytes of data daily.
- Risk is multidimensional. Portfolio optimization and contagion modeling require solving complex combinatorial problems, often beyond the limits of classical hardware.
- Fraud evolves quickly. Detecting new fraud patterns demands faster, adaptive machine-learning models.
- Regulation demands transparency. Compliance checks now involve deeper simulations and audits, putting pressure on computational resources.
In other words, finance lives at the edge of computational possibility, making it an ideal testing ground for both quantum pilots and HPC-powered AI analytics.
Quantum Computing in Fintech: Use Cases and Pilots
1. Fraud Detection:
Fraudulent transactions are one of the hardest problems to tackle on a scale because of the need to minimize false negatives (fraud that slips through) without overwhelming firms with false alerts.
- Intesa Sanpaolo (Italy) tested IBM’s quantum tools and found a variational quantum classifier (VQC) that outperformed classical models with fewer features.
- Deloitte Italy partnered with AWS Braket to build a hybrid quantum neural network that demonstrated superior accuracy in identifying fraudulent transactions compared to traditional ML.
- Research at the UK’s National Quantum Centre showed quantum Restricted Boltzmann Machines could achieve zero false negatives on credit card data sets – a tantalizing result for fraud teams.
These examples suggest that quantum machine learning (QML) could be a game changer in reducing fraud while keeping customer experience smooth.
2. Portfolio Optimization:
Finding the optimal mix of assets for maximum return and minimum risk is a combinatorial nightmare. This is exactly the type of problem where quantum (and quantum-inspired) methods excel.
- Kutxabank (Spain) collaborated with Fujitsu’s Digital Annealer (a quantum-inspired chip) and achieved more stable asset allocations than with classical benchmarks.
- BBVA/Bankia (Spain), working with D-Wave and Multiverse Computing, reported a portfolio with 15% risk and 60% ROI – a dramatic improvement over random allocation. The solution achieved a Sharpe ratio of 12.16 in about 171 seconds, where classical solvers struggled.
- Citi Innovation Labs is partnering with Classiq and AWS Braket to explore QAOA-based (Quantum Approximate Optimization Algorithm) portfolio optimization, aiming for more efficient allocation across client portfolios.
These pilots show how quantum can provide richer portfolio insights and identify opportunities faster than classical methods.
3. Risk Modeling & Forecasting:
Risk contagion and stress forecasting require analyzing network effects across thousands of variables.
- Yapı Kredi (Turkey) used D-Wave’s quantum annealer to analyze its SME borrower network. A computation that “would traditionally take years” was done in seven seconds, enabling early detection of contagion points.
- Goldman Sachs and QC Ware developed a quantum Monte Carlo algorithm for derivatives pricing and risk analysis. This was later demonstrated on IonQ’strapped-ion hardware, marking a significant step toward practical risk simulations.
- JPMorgan Chase, in collaboration with QC Ware, tested quantum deep learning for hedging strategies. The result: quantum deep learning models trained more efficiently than classical equivalents.
These experiments point to a future where quantum enables real-time risk assessment, something classical models often cannot deliver at scale.
Supercomputing in Fintech: The Here and Now
While quantum pilots capture headlines, supercomputing (HPC) is already transforming fintech today. With GPU clusters and accelerated compute, banks are achieving breakthroughs in real-time analytics.
1. High-Performance Risk Analysis
- Modern Monte Carlo simulations for derivatives and stress testing can involve billions of calculations.
- A Dell/NVIDIA HPC system with 8× H100 GPUs set new records on the industry-standard STAC-A2 benchmark, simulating hundreds of millions of option paths in minutes.
- These capabilities mean faster reporting, better scenario analysis, and quicker regulatory compliance.
2. AI-Driven Fraud & Analytics
- BNY Mellon became the first major bank to deploy NVIDIA’s DGX SuperPOD, a supercomputing system with dozens of H100 GPUs.
- The bank uses it for predictive analytics, including fraud detection, deposit forecasting, and trade anomaly detection, therefore reinforcing its ability to “manage, move and keep client assets safe.”
- Other firms use GPU-powered AI to monitor credit card transactions in real time or triage customer interactions efficiently.
3. Real-Time Decisioning & Compliance
- Supercomputers allow firms to crunch massive datasets in milliseconds, powering high-frequency trading and real-time AML/KYC compliance.
- Vendors like IBM, HPE, and Dell provide turnkey HPC solutions tuned for finance, while cloud platforms (AWS, Azure, Google) make HPC accessible on demand.
In contrast to quantum’s exploratory pilots, HPC is a proven, deployed technology, providing immediate returns.
Leading Vendors & Collaborations
- IBM: Offers quantum computing platform, IBM Q, partnered with JPMorgan Chase, HSBC, and others.
- D-Wave: Pioneer in quantum annealing. Used by BBVA/Bankia (Spain) and Yapı Kredi (Turkey) for portfolio and risk pilots.
- AWS Braket: Cloud quantum service. Used by Deloitte, Citi, and Goldman Sachs for fraud and portfolio optimization pilots.
- QC Ware & IonQ: Developed and tested quantum Monte Carlo and quantum deep hedging models with JPMorgan and Goldman.
- Fujitsu: Digital Annealer applied to portfolio optimization with Kutxabank.
- NVIDIA & Dell: Provide HPC infrastructure; BNY Mellon’s DGX SuperPOD is the flagship deployment.
- HPE: Delivers Cray-based HPC+AI platforms to financial institutions.
This ecosystem demonstrates the hybrid future: cloud access to quantum pilots, on-premises supercomputers for daily workloads, and cross-vendor collaborations to bridge theory and practice.
Outlook: From Experimentation to Competitive Advantage
It is easy to dismiss quantum pilots as “experiments,” but the reality is that finance is quietly building expertise. The gains are already tangible: better fraud detection rates, higher Sharpe ratios, faster contagion analysis, and more efficient Monte Carlo simulations.
Meanwhile, HPC continues to evolve as a mainstream force multiplier, enabling firms to keep pace with the rapid pace of markets and regulations.
Together, quantum and supercomputing are not about replacing classical systems but augmenting them:
- Quantum pilots point toward long-term breakthroughs in optimization and forecasting.
- HPC deployments provide immediate scalability and power for AI, risk, and compliance.
For fintech leaders and bank executives, the message is clear: investing in these technologies today builds the muscle memory to compete tomorrow. As financial complexity grows, the winners will be those who can compute faster, analyze deeper, and act sooner.