1.75 quadrillion
The number of floating point operations that the Quriosity supercomputer can perform each second, is equivalent to the computing power of around
50,000 laptops.
China (Mainland)
Meet Quriosity
When it comes to expediting digitalization at BASF,
research and development play a key role. For example, to calculate the most promising polymer structure from thousands of possibilities, you need a high-performance computer with above-average computing power – just a supercomputer like our Quriosity.
With a computing power of 1.75 petaflops (1 petaflops equals one quadrillion floating point operations per second), the new computer, called Quriosity following an online naming contest among employees, offers around 10 times the overall computing power previously available to BASF researchers. It took roughly one year from initial internal planning meetings to start-up.
Installing the supercomputer in Ludwigshafen involved reinforcing the floor in the server room, laying more than 1,000 network cables with a total length of 15 km as well as adding a separate water-cooling system, which can cool the supercomputer with 60,000 liters of water per hour. At full capacity, the supercomputer has an electricity consumption of roughly 600 kilowatts, generating significant waste heat in the process.
Find out more about supercomputers and quantum computers:
With Quriosity, much more complex models are possible, in which significantly more parameters can be varied. This could not only result in substantially reduced development times, but also previously unknown relationships can be recognized and used to advance completely new research approaches.
Find out more about the topics Quriosity is working on:
One application example of our new supercomputer are molecular simulations for detergent formulations. Using Quriosity’s enormous computing power, we are able to explain with these simulations at molecular level how BASF’s existing and potential new detergent ingredients work. This results in a better understanding of products, formulations and processes, and thus enables greater innovation to be achieved in a shorter period of time. Further information regarding BASF’s solutions for Home Care and Industrial & Institutional cleaning can be found here.
Loading ...
The video shows the molecular simulation of a detergent formulation. All detergent ingredients (soap = white blobs, surfactants = greenish blobs, polymers = little sticks) are simulated as well as dirt (red) and the cotton surface (white, at the bottom). At the end of the simulation with about 26 million atoms, there is almost no dirt on the cotton surface. The polymers stabilize the dirt together with the surfactants and inhibit the re-deposition of dirt on the cotton surface.
The performance of molecules is determined by their properties. With the help of Quriosity, we create a huge database with computed molecular properties. These properties can be linked to the performance of BASF products by use of machine learning. Using this approach, our database enables us to find promising molecules for innovative BASF products.
Surfactants are key ingredients in many products for daily use such as dishwashing detergents or skin creams. Even at very low concentration, surfactants can form structures that influence the performance of a product such as the foaming of dishwashing detergents or how easy a cream can be applied to the skin. Using Quriosity we can now study surfactant systems to a high degree of accuracy. With these simulations we support the research and development teams to develop new formulations for home care, industrial & institutional cleaning and personal care products.
Loading ...
Related content:
Digitalization gives researchers additional opportunities to implement their creative ideas and to collaborate intensively with others around the world. In the view of BASF experts, it is essential to integrate digital technologies directly into the daily work of the R&D units. Direct access to knowledge-based systems is necessary to enable effective problem-solving and it opens up new horizons. A cloud-based app platform, for example, will make it considerably easier for all researchers to expand knowledge networks. In recent months, successful projects by BASF researchers have already demonstrated the enormous potential that digitalization offers in research. For example, researchers were able for the first time to conduct a systematic investigation of the data on catalysts used in the production of the intermediate product ethylene oxide. The investigation found correlations between the formulations and the application properties of the catalysts, which enabled their performance and lifetime to be predicted more accurately and faster. Another example are computer simulations of microencapsulation. It is used to protect active ingredients, for example from moisture and oxygen. Simulations help to better understand and predict the complex chemical and physical interactions within the microcapsule, allowing laboratory test series to be planned more precisely.
BASF is also playing an important role in the digital transformation of acgriculture, where it relies on internal and external cooperation. The online application Maglis® helps farmers use available information more efficiently and make better decisions regarding the cultivation of their fields. The firm ZedX, which BASF acquired at the end of May, specializes in developing agronomic models for weather, plant growth and infestation with diseases, weeds and pests. BASF and ZedX have already jointly developed a model that, based on weather and environmental conditions, identifies the right window of application for a BASF herbicide.