Registration does not begin until 02.10.2026 at 08:00. It ends on 12.10.2026 at 23:59.

The registration is restricted to the following fields or courses of study:

• Leuphana Bachelor / Major Wirtschaftsinformatik (ab Studienbeginn WiSe 20/21) / alle Semester
• Leuphana Bachelor / Major Wirtschaftsinformatik (bis Studienbeginn WiSe 19/20) / alle Semester
• Leuphana Bachelor / Major Studium Individuale / alle Semester
• Leuphana Bachelor / Minor Studium Individuale / alle Semester

Course content:
This course introduces Business Informatics students to GPU programming and its applications in business data processing and analytics. While CPUs are the traditional workhorses of computing, many business tasks—such as large-scale data analysis, financial modeling, and real-time simulations—can benefit immensely from the parallel processing power of GPUs.
Students will begin with the fundamentals of C++ programming and then transition into GPU programming using CUDA. The course culminates in exploring the architecture of GPUs and how to optimize programs for maximum efficiency. Each module ends with a project focused on real-world business applications, providing students with practical experience that ties directly into their field.
Course Modules
1. C++ Fundamentals for Business Applications
Students will learn how to write, compile, and execute basic C++ programs, tailored for data-centric tasks. Topics include syntax, variables, control flow, functions, standard containers, memory management, and file I/O. The focus is on processing business-related data efficiently using C++.
Project:
Build a program that reads structured business data (e.g., customer transactions, sales records), performs statistical computations, and outputs results for reporting.
2. GPU Programming with CUDA
This module introduces CUDA and the core principles of GPU programming. Students will learn how to write CUDA kernels, manage memory, and understand the structure of threads and blocks. By transitioning an existing CPU-based program into CUDA, students will directly experience the performance benefits.
Project:
Implement a GPU-accelerated program that performs heavy data computations (e.g., real-time financial time-series analysis or batch processing of customer data), and compare performance with the CPU version.
3. Advanced GPU Techniques
In this module, students will explore key parallel programming techniques that are foundational for high-performance data analysis:
Reduction: Used for aggregating data (e.g., total sales, max transaction value).
Scan (Prefix Sum): Useful for tasks like cumulative financial calculations or indexing grouped data.
Histogram: Ideal for categorizing data distributions (e.g., customer segmentation, sales frequency analysis).
Students will learn how to implement these operators using CUDA, understand their parallel execution models, and see how they apply to real-world business datasets.
Project:
In this project, students will develop a CUDA-accelerated analytics tool that processes a large dataset of business transactions (e.g., sales, customer activity, or stock prices) and computes:
Total and maximum transaction values using reduction
Daily or weekly cumulative revenue using scan
Frequency of sales by category or customer segment using a histogram

By the end of the course, students will be able to:
Write and debug C++ code for business data processing.
Leverage CUDA to write GPU-accelerated programs for business applications.
Understand GPU architecture to optimize data processing performance.
Evaluate and apply GPU solutions for real-world Business Informatics challenges.

Further information on teaching evaluation: https://www.leuphana.de/en/teaching/quality-management/evaluation/course-evaluation.html