Global FMCG companies are turning to AI and HPC to streamline product innovation and stay ahead of shifting consumer tastes
By Ashley Woodbridge, CTO, Lenovo, META
When thinking about how crisps are made, your mind likely goes to factory lines and seasoning machines, not HPC (high-performance computing) clusters. But behind the flavors we enjoy, is a growing wave of technological innovation. HPC, combined with AI, is transforming the way food is developed, helping brands design new tastes more quickly, more efficiently and with a lighter environmental footprint.
This combination of computing power and AI is opening up a new era of flavor development. The speed of AI-supported digital simulations has boosted the traditional trial-and-error process, making it possible to test more ideas with less waste in less time. Whether it’s creating a tangy new crisp or improving texture in baked goods, technology is now a key ingredient in food innovation.
Lenovo’s HPC (High Performance Computing) solutions are revolutionizing food innovation in the Middle East, enabling companies to develop new flavors and products more efficiently, intelligently and with a reduced environmental impact. By leveraging Lenovo’s advanced HPC systems, food manufacturers can conduct complex simulations to examine molecular interactions, refine ingredient combinations, and anticipate taste profiles. All of this can now be done without the need for resource-heavy physical trials. This approach not only speeds up product development but also supports the region’s sustainability initiatives. With Lenovo – food innovators are enhancing culinary creativity while minimizing waste and boosting efficiency.
From Assembly Lines to Algorithms
HPC has long been used across industries to model and optimise complex systems, including aerodynamics in motorsports and molecular structures in drug development. In the food industry, it has traditionally helped improve packaging performance and production line efficiency. Today, its role is expanding. When paired with AI solvers, a type of software that uses AI to learn from data and automatically find solutions to complex problems, HPC allows companies to simulate and refine how ingredients interact with each other with significantly greater precision and speed.
These AI solvers are making a significant difference by tackling one of food science’s most intricate challenges: fluid dynamics. Many ingredients used in food processing, such as oils, flavourings and emulsions, behave like complex fluids. Simulating how they blend or react under different conditions requires computational fluid dynamics (CFD), a highly detailed and time-consuming process that uses complex equations to work out how liquids behave in various settings. AI solvers accelerate this work by learning from past simulations and predicting the outcome of new ones without needing to compute every step in between.
Why This Matters for Food Innovation
In the past, adjusting a single ingredient in a recipe required re-running an entire CFD simulation. Because these are difficult to run in parallel, the process was slow and costly. Even a small change, like switching an ingredient supplier or adapting a recipe for a new market, meant starting over from scratch.
AI solvers change this approach completely. By using machine learning models trained on a few representative simulations, solvers can estimate the final result without performing every calculation. This ability to “skip to the end” allows developers to test many more scenarios in a shorter period.
The result is a more agile and creative product development process. Food makers can quickly prototype new flavours, fine-tune texture or consistency, and adjust for regional taste preferences. What once took weeks or even months can now be done in days.
A More Sustainable Approach
Beyond speed, these tools also bring sustainability benefits. Full CFD simulations consume significant amounts of energy due to their reliance on large-scale HPC infrastructure. By reducing the need for such extensive computing resources, AI solvers help lower the environmental impact of product development.
Many solvers are built using lightweight AI architectures, such as multi-layer perceptrons, that can operate on a single graphics processing unit (GPU). These models are far less energy-intensive than traditional HPC workloads. By enabling researchers to experiment with digital prototypes rather than physical ones, this approach conserves both time and resources.
Putting AI and HPC to Work
This technology is already having a real impact. Global FMCG companies are turning to AI and HPC to streamline product innovation and stay ahead of shifting consumer tastes. Mondelez International, for example, has used AI to speed up development for over 70 products, including Gluten-Free Golden Oreos. Kellanova (formerly Kellogg Company) is using CFD simulations to model flavor chemistry for its Pringles range, while AI solvers are helping to run those simulations more quickly.
Firms such as AI Palette, which specialize in predictive analytics for food innovation, are enabling producers to simulate how ingredients will behave together and anticipate which combinations are likely to appeal to consumers. These simulations allow food developers to refine recipes without waiting for physical testing results.
What’s Next for the Industry?
As AI and HPC continue to advance, their influence on the food industry will only grow. With more powerful solvers and better data, companies will be able to deliver personalized flavors, optimise formulations for dietary needs and shorten time to market for new products.
This acceleration also means a lower barrier to innovation. Food makers can take creative risks more confidently, knowing that the development process is more responsive and cost-effective. In turn, shoppers can expect to see a greater variety of products on store shelves, with new flavors and formats appearing more frequently.
The Crunch Behind the Crunch
Next time you reach for a packet of crisps or try a new snack, consider what went into it. Behind the familiar flavour could be layers of digital modelling, AI-driven predictions and high-speed calculations. While the process may be invisible, its results are anything but. AI and HPC are quietly reshaping how food is created, one crunch at a time.
