IonQ Unveils The Power Of Its Next-Generation Quantum Computer Along With Quantum Finance Announcements


This week, IonQ, Inc. (IonQ) announced the research results for two separate finance-related quantum research projects. The announcement covered one with Fidelity Center for Applied Technology, and one with Goldman Sachs and QC Ware. 

While classical computers use bits for computation, quantum computers use quantum bits or “qubits.” IonQ uses ion qubits created using precision lasers to remove an outer electron from an atom of ytterbium. IonQ has plans to evolve its existing architecture to a more advanced version sometime in the future. The power of its new hardware was demonstrated in the Goldman Sachs and QC-Ware research below. Moor Insights & Strategy previously wrote about IonQ’s evolution to its new architecture here

Even though the finance industry is computationally intensive, applications containing large numbers of variables are too complex to perform on classical computers. Eventually, solutions for classically intractable problems will become available using quantum computers. Most experts believe it will likely take another five to seven years before today’s quantum machines have enough power to move from experimental prototypes to production environments.  

When that happens, financial institutions will begin to use quantum computers for everything from pricing to option derivatives to risk management to liquidity coverage. Even though that is still a few years away, most major finance companies have already begun to staff quantum computing research departments. 

Here is a summary of IonQ’s recent announcements:

IonQ and Fidelity Center for Applied Technology (FCAT)


FCAT and IonQ researchers used IonQ’s cloud-based quantum computer to develop a quantum machine learning (QML) proof of concept that achieved far better results than previous research.  It’s important to note that this research also demonstrated that quantum computers can outperform classical computers for limited price correlation analysis in the finance industry. Technical details of the study are available here.

Historical data is heavily used for training and analysis in today’s financial models. For the output to be correct and free from bias, the training data must accurately reflect the characteristics of the modeled scenario.  A standard testing process called “backtesting” uses data separate from the training data but believed to be similar can determine if a model produces accurate results.  However, backtesting is often insufficient because it is challenging to obtain test data that accurately depicts all the market scenarios represented in an extensive training dataset.

The FCAT-IonQ team built a quantum AI model that created a new and more accurate set of synthetic data to obtain accurate data for backtesting. The synthetic data was created from samples of the same data used to train the model. This procedure is much like the uncanny ability of AI models trained on facial images to create new and authentic faces that look identical to real people. 

Instead of facial images, the IonQ and FCAT teams modeled numerical relationships contained in the daily returns of Apple and Microsoft stock from 2010 to 2018. Two quantum machine learning algorithms used this data to produce a highly accurate synthetic data set for backtesting.

A few considerations:

  • Backtesting is an important data science technique and the ability to create accurate synthetic data is an important achievement.
  • The Quantum Machine Learning Model in this research only involved two stocks. To be truly useful, the model must be significantly scaled up to accommodate a greater number of stocks.  Increasing the number of stocks makes computation increasingly more complex, requiring quantum computers with thousands of qubits.
  • The ability of classical computers to create synthetic data for stock correlations is limited because real-world use cases require too many variables with complex dependencies.
  • This research demonstrates that generative learning algorithms on trapped ion quantum computers with a small number of qubits can outperform equivalent classical generative learning.

IonQ, Goldman Sachs, and QC-Ware


Using IonQ’s newest quantum computing hardware, Goldman Sachs and QC-Ware teamed up to push quantum boundaries beyond previous research. The team demonstrated a quantum algorithm developed by QC-Ware for Monte Carlo simulations on IonQ’s recently announced quantum processing unit (QPU), Evaporated Glass Trap. Applications of quantum Monte Carlo methods to problems in computational finance have been the subject of several previous research papers. That research involved applying quantum Monte Carlo to specific financial problems such as pricing simple options and credit risk calculations.

According to IonQ, its new QPU has an order of magnitude better fidelity and better throughput than its current generation of quantum processors. In its press release, Peter Chapman, CEO and President of IonQ, emphasized the importance of using a combination of state-of-the-art hardware and best-in-class quantum algorithms. 

In the published results, the quantum researchers also attributed the project’s success to the high fidelity of IonQ’s quantum hardware. The researchers also stated that similar experiments were attempted using other quantum hardware available on the cloud but obtained “considerably worse results.”

Quantum computers are expected to not only have a major impact for Monte Carlo simulations, but in other areas of science and engineering as well. Monte Carlo simulations demonstrated by this research are especially important to finance in the areas of risk and derivative pricing for such things as options.  Some estimates size the derivatives market to be worth over one quadrillion dollars. Monte Carlo simulations are usually run on classical computers and require the algorithm to be run a number of times to obtain an estimated answer with acceptable precision. When large fault-tolerant quantum computers become available, it will significantly reduce the amount of time needed to obtain solutions for complicated Monte Carlo problems containing a large number of variables. The precision of estimated answers can be improved by increasing the number of samples. For example, to increase a classical computer’s answer precision by one order of magnitude requires increasing sampling by 100X.  For an equivalent accuracy, a quantum computer would only require a sampling increase of 10X.  In finance, time is an important commodity. A few seconds in a large, fast-moving market such as stocks and options can mean the difference between a profit or a loss.


  • The research paper for this announcement is published here .
  • Although IonQ’s new generation QPU was used in this research, it is not generally available. 
  • Moor Insights & Strategy believes that the finance domain will be one of the first to use production quantum applications.

Note: Moor Insights & Strategy writers and editors may have contributed to this article.