SAS’s case for “quantum AI” is chiefly about using quantum annealing to tackle selected optimization problems, not replacing enterprise AI or conventional computing. At SAS Innovate 2025, Procter & Gamble (P&G) described a constrained manufacturing-mixing problem where a hybrid quantum-and-classical approach reportedly cut solution time from six hours to 12 minutes. That is a promising demonstration, not proof that quantum methods are faster or better for business workloads in general.
What SAS means by “quantum AI”
In ITPro’s May 8, 2025 account, “quantum AI” refers mainly to quantum annealing applied to optimization: searching for good solutions among many combinations while satisfying constraints. That is narrower than general-purpose quantum computing and distinct from a claim that quantum systems can replace machine learning, analytics, or enterprise AI.
The practical question is therefore not whether quantum is universally superior, but whether a particular workload has a structure that suits the method. SAS COO Gavin Day made a similar point by comparing it with GPUs: “There are some instructions and problems that GPUs are excellent at solving, there are others that actually are slower and worse.”
What happened in the P&G manufacturing example
At SAS Innovate 2025, P&G director of product and innovation Krista Comstock described a mixing-tank optimization problem. The task involved selecting ingredient combinations while observing constraints intended to prevent cross-contamination. ITPro reports Comstock’s estimate of 10114 possible ingredient mixes; that figure is an attributed estimate from the event account, not an independently verified count.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problems#1 Best Overall
| Approach | Reported time | Reported result and qualification |
|---|---|---|
| Traditional SAS Viya solver algorithms | Six hours | Conference demonstration timing as reported by ITPro; benchmark setup and reproducibility details were not published in the account. |
| Quantum annealing | Two minutes | Reportedly produced unreliable results at scale in this example. |
| Hybrid quantum and classical | 12 minutes | Quantum handled most of the problem and traditional solvers performed final calculations; Comstock described this as a 30-fold reduction compared with the traditional method. |
The reported hybrid result is the most relevant comparison because it includes the classical work needed to finish the calculation. Still, ITPro’s conference account does not provide hardware details, a full benchmark setup, or reproducibility data, and no primary P&G or SAS technical report was surfaced to independently verify the figures. The timings should be treated as a demonstration claim about this workload, not a forecast for other factories, optimization tasks, or enterprise systems.
Why a hybrid workflow matters
The example’s quantum-only result was reportedly unreliable at scale, so the described workflow paired quantum processing with conventional solvers. That illustrates a plausible enterprise pattern: use a specialized method for part of a problem, then rely on classical tools where they are more dependable or appropriate.
Rank #2
It also means “quantum speedup” can be misleading if it refers only to a quantum substep. For a useful comparison, organizations need the end-to-end time and solution quality of the complete workflow, including classical post-processing and the effort required to prepare the problem.
What the adoption figures do—and do not—show
ITPro reported results from a 2025 SAS survey of 500 business leaders worldwide. More than 60% said they were investing in or investigating quantum AI’s potential for their organizations, while 38% expressed concern about its costs. Those are SAS survey findings as reported by ITPro, not independently verified market-wide adoption rates. The account does not provide enough methodology to characterize the sample beyond its reported size and worldwide scope.
Recommended Free Tools
Cost is only one practical hurdle. ITPro’s account also points to specialized hardware, algorithm maturity, and continuing research and development as adoption barriers. SAS COO Gavin Day said technology providers need to make enterprise-scale projects possible while lowering the barrier to entry for mid-sized and smaller businesses.
How an enterprise should assess a quantum pilot
A pilot should test whether a method improves a defined business problem against credible alternatives, not simply whether quantum hardware can be accessed. Compare the full workflow on the same problem and constraints, and set success criteria before running it.
- Workload fit: Identify a concrete optimization problem with meaningful combinations and constraints. Do not assume that an ordinary analytics or machine-learning task will benefit from quantum annealing.
- Solution quality and reliability: Measure whether the method repeatedly produces usable solutions, including at the scale the business actually needs.
- End-to-end time: Include problem preparation, quantum processing, classical solver work, and final calculations—not just the quantum portion.
- Total cost and access: Account for specialized hardware access, software, expertise, and the ongoing effort needed to maintain the workflow.
- Fair baseline: Compare against relevant conventional solvers using the same constraints and quality requirements. A dramatic timing claim is hard to assess without the benchmark setup and repeatability information.
SAS principal product manager for Quantum Computing Amy Stout described the company’s goal as making quantum “simple, fast, and intuitive” for customers. That is an aspiration, not evidence that enterprise deployment is already straightforward; a pilot should establish operational practicality as well as technical performance.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What SAS’s claims support
The P&G example supports a limited but useful conclusion: a hybrid quantum-and-classical workflow may be worth exploring for particular constrained optimization problems. It does not establish that quantum computing broadly outperforms conventional computing, that quantum annealing is reliable for every large problem, or that the reported timings will transfer to another workload.
Free tools Windows power users keep installed
One-click scans. No signup required.
Best Value
ITPro’s 2025 account also named D-Wave Quantum Inc., IBM, and QuEra Computing as companies SAS was working with. That report establishes only what it said at that time; it does not establish current partnership status, scope, or product availability.
Source: ITPro, “SAS thinks quantum AI has huge enterprise potential – here’s why,” May 8, 2025.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
