Research Hub > For Financial Services Firms, HPC Is Money Well Spent
3 min

For Financial Services Firms, HPC Is Money Well Spent

High-performance computing is a powerful, proven way to gain fresh insights about what’s happening and what will happen next.

CDW Expert CDW Expert

How will Apple’s stock price react when the next iPhone is announced? Or when the next Samsung Galaxy and Google Pixel are unveiled? What about if there’s a software update that creates issues with the iPhone or the Apple Watch?

Until someone invents a crystal ball, high-performance computing (HPC) is the only proven way to predict how those kinds of events may unfold. That’s why investment management firms, banks and other financial services organizations are increasingly using HPC to make more informed decisions about:

  • Prepayment and default risks on bonds
  • What social media posts portend about sales of a forthcoming product
  • How quickly and how long weather events or political regime changes may affect certain stock prices
  • What a FICO score, ZIP code or loan-to-value ratio indicates about a borrower’s likelihood of defaulting on a mortgage or paying it off early

Boiled down, HPC is about having a lot of nodes analyze a lot of data to get deep, actionable insights much faster than other techniques and technologies can deliver. How fast? Before joining CDW, I worked in the energy sector, where a seismic analysis often took six months. When we switched to HPC, that time frame for the same analysis dropped to six days.

Better Insights for Better Decisions

In financial services, HPC is ideal for analyzing historical data to better understand what’s happening or what’s about to happen — especially when you pair HPC with artificial intelligence (AI) and machine learning (ML). For example, combining those three capabilities can help banks minimize churn, fraud and customer support costs by analyzing purchases made on the debit and credit cards they issue.

Suppose a Michigan customer’s card is suddenly used in Cancun. Is that fraud? HPC, AI and ML could determine that the risk is low because the customer used that card to buy a ticket to Cancun a few months earlier. Without that instant insight, her card would be declined. That would cost the bank in several ways, including the time a live agent spends responding to her irate call from Mexico, lost business if she churns as soon as she gets back and more lost business if her social network hears about it and some decide to bank elsewhere.

Refine Your HPC Strategy

As with any other technology, the ROI on HPC depends heavily on how it’s implemented. For instance, take a holistic view of the data, from ingest to visualization. Does the data have to be normalized, deduplicated or both to get it in a usable format? Will there be disparate types of data, such as hot or cached, alongside archived? Will accelerators be required to support real-time analytics, or will the work be done in batches? These and other considerations affect choices such as storage, file systems and whether the cloud has a role.

Security is another consideration. Businesses often use HPC to analyze confidential data, such as customer transaction histories. Make sure that information is encrypted at rest and in flight, such as when the HPC environment extends to the cloud.

That’s a lot to think about, which is another example of how HPC is like most other advanced technologies: Choosing the right partner is critical for identifying all of your options and selecting the one that best supports your business goals.

Visit or or contact your account manager to learn more.

This blog post brought to you by: