The Paderborn Center for Parallel Computing stands at the forefront of advancing high-performance computing technologies and methodologies. We are driven by a commitment to innovation and excellence at the confluence of three pivotal areas:

1. Innovative computer architectures,
2. Co-design of efficient methods for large-scale scientific simulation,
3. Advanced applications in computational science and engineering.

Our overarching goal is to pioneer "Efficient and Scalable High-Performance Computing," aiming to shape the way computational science and engineering problems are approached and solved.

We are dedicated to pushing the frontiers of HPC through rigorous research in these three strategic areas, laying the groundwork for future scientific breakthroughs and their practical applications.

Our institute is a beacon in the field of application-specific and domain-specific computing, boasting profound expertise and accomplishments. We design and develop highly customized and optimized processing architectures with a strong emphasis on Field-Programmable Gate Arrays (FPGAs). These innovations result in significant improvements in performance and energy efficiency. Beyond FPGAs, our research encompasses a variety of massively parallel computing architectures and advanced networking technologies. In addition to these forward-looking studies of FPGAs, we are also exploring the potential of other massively parallel computing architectures, including Graphics Processing Units (GPUs) and multi/many-core CPUs, which are the backbone of today’s HPC systems but continue to offer numerous research opportunities.

As a national HPC center, we are committed to transitioning our research and technologies from experimental testbeds to production-grade HPC systems. This enables us to evaluate our contributions on a larger scale with real-world applications. Our unique FPGA installations are among the most advanced in the world, underscoring our role in integrating and leveraging innovative technologies such as optical switches for next-generation computing solutions.

Recognizing that simulation methods and computer architectures are interdependent, we are taking a holistic approach to improve simulation efficiency. By tailoring simulation methods to the unique characteristics of targeted hardware, we unlock unprecedented efficiency potential. This effort extends to a true co-design of application/simulation methods and computer architectures, especially when using FPGAs. This integrated approach requires a broad spectrum of expertise, spanning applications, methods, and architectures, and fosters a fruitful collaboration between HPC specialists, application developers, and domain experts.

Our research in applications focuses on tackling the toughest challenges in computational physics and chemistry. These problems often emerge from our consulting and support services for users of our HPC infrastructure, leading to collaborative research and development efforts.

By working closely with our user community, we not only address specific computational challenges, but also employ these complex applications to showcase our methodological and architectural innovations. This hands-on approach ensures that our research contributions are not only theoretically sound, but also directly applicable, thereby advancing the fields of computational science and engineering.

We are actively committed to open research and generally publish our research results as open source on GitHub (PC2 Organization at GitHub) or integrate our developments into existing community software, such as CP2K or Julia.