Christian Schwarz |
||
 |
Phone Visitor's Address |
christian.schwarz1@tu-dresden.de +49 (0)351 463 43732 Helmholtzstrasse 18, BAR III62 |
Christian Schwarz obtained his Bachelor's and Master's degree in Computer Science from the TU Dresden in 2021 and 2025, respectively. He wrote his master thesis with Prof. Horst Schirmeier at the Chair of Operating Systems on “Scheduling Heterogeneous Memory for Database Query Execution”. Christian is an experienced software developer in areas that span databases, operating systems and compilers.
In November 2025, he joined the chair as research assistant, where he works on Machine Learning for Compilers,
building on prior work from our group by Alexander Brauckmann and Anderson da Silva, and recent joint work with Fernando M Q Pereira. Concretely, he works on the TransT DFG project on Cross-domain, Cross-platform Transfer Learning and Auto-tuning.
2026
- Anderson Faustino da Silva, Christian Schwarz, Jeronimo Castrillon, "Structure-Aware Compiler Auto-Tuning", In Proceeding: Brazilian Symposium on Programming Languages (SBLP), Sep 2026. [Bibtex & Downloads]
Structure-Aware Compiler Auto-Tuning
×Reference
Anderson Faustino da Silva, Christian Schwarz, Jeronimo Castrillon, "Structure-Aware Compiler Auto-Tuning", In Proceeding: Brazilian Symposium on Programming Languages (SBLP), Sep 2026.
Abstract
Modern compilers such as GCC expose large sets of optimization
flags whose interactions induce a high-dimensional configuration space, making manual tuning impractical and exhaustive auto-tuning infeasible. Recent approaches, including Bayesian-optimization-based search and critical-flag selection via static analysis, improve sample efficiency but can exhibit premature convergence to locally
optimal configurations. We present SACT, a structure-aware compiler auto-tuning framework that integrates static program characterization with surrogate-guided search to reduce stagnation. SACT extracts lightweight structural features, initializes a program embedding, and uses surrogate-sensitivity-based relevance estimates to adaptively focus the search on influential flags while maintaining controlled exploration. The embedding is refined online from compile–execute measurements, enabling the surrogate to capture program-specific flag interactions over the course of tuning. Experiments on GCC 15.2.0 using the PolyBench and MiBench benchmark
suites show that SACT achieves a better geometric average speedup than prior methods while requiring a low wall-clock budget.Bibtex
@inproceedings{silva_sblp26,
title={Structure-Aware Compiler Auto-Tuning},
author={Anderson Faustino da Silva and Christian Schwarz and Jeronimo Castrillon},
booktitle={Brazilian Symposium on Programming Languages (SBLP)},
location = {São Paulo, Brazil},
year={2026},
month = sep,
abstract = {Modern compilers such as GCC expose large sets of optimization
flags whose interactions induce a high-dimensional configuration space, making manual tuning impractical and exhaustive auto-tuning infeasible. Recent approaches, including Bayesian-optimization-based search and critical-flag selection via static analysis, improve sample efficiency but can exhibit premature convergence to locally
optimal configurations. We present SACT, a structure-aware compiler auto-tuning framework that integrates static program characterization with surrogate-guided search to reduce stagnation. SACT extracts lightweight structural features, initializes a program embedding, and uses surrogate-sensitivity-based relevance estimates to adaptively focus the search on influential flags while maintaining controlled exploration. The embedding is refined online from compile–execute measurements, enabling the surrogate to capture program-specific flag interactions over the course of tuning. Experiments on GCC 15.2.0 using the PolyBench and MiBench benchmark
suites show that SACT achieves a better geometric average speedup than prior methods while requiring a low wall-clock budget.},
numpages = {9},
organization={SBC}
}Downloads
No Downloads available for this publication
Permalink
2022
- Christian Schwarz, Viktor Reusch, Maksym Planeta, "DMA Security in the Presence of IOMMUs", Gesellschaft für Informatik e.V., 2022. [doi] [Bibtex & Downloads]
×
Reference
Christian Schwarz, Viktor Reusch, Maksym Planeta, "DMA Security in the Presence of IOMMUs", Gesellschaft für Informatik e.V., 2022. [doi]
Bibtex
@article{https://doi.org/10.18420/fgbs2022f-04,
doi = {10.18420/FGBS2022F-04},
url = {http://dl.gi.de/handle/20.500.12116/38507},
author = {Schwarz, Christian and Reusch, Viktor and Planeta, Maksym},
keywords = {IOMMU, DMA, PCIe, SR-IOV, RDMA, Thunderbolt, Rowhammer},
language = {en},
title = {DMA Security in the Presence of IOMMUs},
publisher = {Gesellschaft für Informatik e.V.},
year = {2022}
}Downloads
No Downloads available for this publication
Permalink
- Johannes Fett, Urs Kober, Christian Schwarz, Dirk Habich, Wolfgang Lehner, "Accelerating Parallel Operation for Compacting Selected Elements on GPUs", Springer International Publishing, pp. 186–200, 2022. [doi] [Bibtex & Downloads]
Accelerating Parallel Operation for Compacting Selected Elements on GPUs
×Reference
Johannes Fett, Urs Kober, Christian Schwarz, Dirk Habich, Wolfgang Lehner, "Accelerating Parallel Operation for Compacting Selected Elements on GPUs", Springer International Publishing, pp. 186–200, 2022. [doi]
Bibtex
@inbook{Fett_2022, title={Accelerating Parallel Operation for Compacting Selected Elements on GPUs}, ISBN={9783031125973}, ISSN={1611-3349}, url={http://dx.doi.org/10.1007/978-3-031-12597-3_12}, DOI={10.1007/978-3-031-12597-3_12}, booktitle={Euro-Par 2022: Parallel Processing}, publisher={Springer International Publishing}, author={Fett, Johannes and Kober, Urs and Schwarz, Christian and Habich, Dirk and Lehner, Wolfgang}, year={2022}, pages={186–200} }Downloads
No Downloads available for this publication
Permalink