Alexander Geiß is a doctoral researcher who recently joined the HRZ in Darmstadt. His research is on performance analysis and modeling of high-performance computing (HPC) applications. In particular, he works on making empirical performance models robust against system noise. His contributions include:

• noise-resilient performance modeling using deep neural networks 
• automatically finding hardware counters that are resilient to noise 
• validating GPU ports of applications with differential performance models 

 His work is closely linked to the HPC lab’s performance-modeling tool Extra-P.