Karl Fürlinger
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Karl Fürlinger has a total of 70 co-authors. Here are the top co-authors by number of shared publications:
Michael Gerndt
Institut für Informatik, Lehrstuhl für Rechnertechnik und Rechnerorganisation, Technische Universität München, Boltzmannstr. 3, DE-85748 Garching
7 shared publicationsErwin Laure
5 shared publicationsShirley Moore
5 shared publications- Dieter Kranzlmüller
Ludwig-Maximilians-Universität München, Munich, Germany
4 shared publications Roger Kowalewski
Ludwig-Maximilians-Universität München
4 shared publicationsXavier Aguilar
4 shared publications
Latest Publications
- Open Access
- 25 January 2026
- Published by Association for Computing Machinery (ACM)
- p. 284-295
Revisiting Communication Software Offloading for MPI+Threads: Reducing Contention and Improving Overlap on Many-Core Systems
- 25 January 2026
- Published by Association for Computing Machinery (ACM)
- p. 284-295
Open AccessRevisiting Communication Software Offloading for MPI+Threads: Reducing Contention and Improving Overlap on Many-Core Systems
- Open Access
- 4 September 2024
- Published by Springer Nature
- p. C1-C2
Correction to: Euro-Par 2023: Parallel Processing Workshops
- 4 September 2024
- Published by Springer Nature
- p. C1-C2
Open AccessCorrection to: Euro-Par 2023: Parallel Processing Workshops
- Open Access
- 12 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 1334-1342
Modelling Data Locality of Sparse Matrix-Vector Multiplication on the A64FX
One of the novel features of the Fujitsu A64FX CPU is the sector cache. This feature enables hardware-supported partitioning of the L1 and L2 caches and allows the programmer control of which partition is used to place data in. This paper performs an in-depth study of how to apply the sector cache to a frequently used sparse matrix-vector multiplication (SpMV) kernel. A performance model based on reuse analysis is used to better understand situations where the sector cache leads to improved reuse and to predict the cache behavior. The model correctly predicts the number of L2 cache misses within 2–3 % for sequential and parallel SpMV with 48 threads using a collection of 490 sparse matrices. Further experiments show the effect of various sector cache configurations on performance. A median speedup of about 1.05 × is achieved, whereas the maximum speedup is about 1.6 ×.
- 12 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 1334-1342
Open AccessModelling Data Locality of Sparse Matrix-Vector Multiplication on the A64FX
One of the novel features of the Fujitsu A64FX CPU is the sector cache. This feature enables hardware-supported partitioning of the L1 and L2 caches and allows the programmer control of which partition is used to place data in. This paper performs an in-depth study of how to apply the sector cache to a frequently used sparse matrix-vector multiplication (SpMV) kernel. A performance model based on reuse analysis is used to better understand situations where the sector cache leads to improved reuse and to predict the cache behavior. The model correctly predicts the number of L2 cache misses within 2–3 % for sequential and parallel SpMV with 48 threads using a collection of 490 sparse matrices. Further experiments show the effect of various sector cache configurations on performance. A median speedup of about 1.05 × is achieved, whereas the maximum speedup is about 1.6 ×.
- Open Access
- 26 June 2023
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 35 (24)
From reactive to proactive load balancing for task‐based parallel applications in distributed memory machines
Summary: Load balancing is often a challenge in task‐parallel applications. The balancing problems are divided into static and dynamic. “Static” means that we have some prior knowledge about load information and perform balancing before execution, while “dynamic” must rely on partial information of the execution status to balance the load at runtime. Conventionally, work stealing is a practical approach used in almost all shared memory systems. In distributed memory systems, the communication overhead can make stealing tasks too late. To improve, people have proposed a reactive approach to relax communication in balancing load. The approach leaves one dedicated thread per process to monitor the queue status and offload tasks reactively from a slow to a fast process. However, reactive decisions might be mistaken in high imbalance cases. First, this article proposes a performance model to analyze reactive balancing behaviors and understand the bound leading to incorrect decisions. Second, we introduce a proactive approach to improve further balancing tasks at runtime. The approach exploits task‐based programming models with a dedicated thread as well, namely . Nevertheless, the main idea is to force not only to monitor load; it will characterize tasks and train load prediction models by online learning. “Proactive” indicates offloading tasks before each execution phase proactively with an appropriate number of tasks at once to a potential victim (denoted by an underloaded/fast process). The experimental results confirm speedup improvements from to in important use cases compared to the previous solutions. Furthermore, this approach can support co‐scheduling tasks across multiple applications.
- 26 June 2023
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 35 (24)
Open AccessFrom reactive to proactive load balancing for task‐based parallel applications in distributed memory machines
Summary: Load balancing is often a challenge in task‐parallel applications. The balancing problems are divided into static and dynamic. “Static” means that we have some prior knowledge about load information and perform balancing before execution, while “dynamic” must rely on partial information of the execution status to balance the load at runtime. Conventionally, work stealing is a practical approach used in almost all shared memory systems. In distributed memory systems, the communication overhead can make stealing tasks too late. To improve, people have proposed a reactive approach to relax communication in balancing load. The approach leaves one dedicated thread per process to monitor the queue status and offload tasks reactively from a slow to a fast process. However, reactive decisions might be mistaken in high imbalance cases. First, this article proposes a performance model to analyze reactive balancing behaviors and understand the bound leading to incorrect decisions. Second, we introduce a proactive approach to improve further balancing tasks at runtime. The approach exploits task‐based programming models with a dedicated thread as well, namely . Nevertheless, the main idea is to force not only to monitor load; it will characterize tasks and train load prediction models by online learning. “Proactive” indicates offloading tasks before each execution phase proactively with an appropriate number of tasks at once to a potential victim (denoted by an underloaded/fast process). The experimental results confirm speedup improvements from to in important use cases compared to the previous solutions. Furthermore, this approach can support co‐scheduling tasks across multiple applications.
- Open Access
- 28 April 2023
- Published by Springer Nature
- p. 263-275
Proactive Task Offloading for Load Balancing in Iterative Applications
Load imbalance is often a challenge for applications in parallel systems. Static cost models and pre-partitioning algorithms distribute the load at the beginning. Nevertheless, dynamic changes during execution or inaccurate cost indicators may lead to imbalance at runtime. Reactive work-stealing strategies can help monitor the execution and perform task migration to balance the load. However, the benefits depend on migration overhead and assumption about future execution.Our proactive approach further improves existing solutions by applying machine learning to online load prediction. Following that, we propose a fully distributed algorithm for adapting the prediction result to guide task offloading. The experiments are performed with an artificial test case and a realistic application named Sam(oa) $$^2$$ on three systems with different communication overhead. Our results confirm improvements for important use cases compared to previous solutions. Furthermore, this approach can support co-scheduling tasks across multiple applications.- 28 April 2023
- Published by Springer Nature
- p. 263-275
Open AccessProactive Task Offloading for Load Balancing in Iterative Applications
Load imbalance is often a challenge for applications in parallel systems. Static cost models and pre-partitioning algorithms distribute the load at the beginning. Nevertheless, dynamic changes during execution or inaccurate cost indicators may lead to imbalance at runtime. Reactive work-stealing strategies can help monitor the execution and perform task migration to balance the load. However, the benefits depend on migration overhead and assumption about future execution.Our proactive approach further improves existing solutions by applying machine learning to online load prediction. Following that, we propose a fully distributed algorithm for adapting the prediction result to guide task offloading. The experiments are performed with an artificial test case and a realistic application named Sam(oa) $$^2$$ on three systems with different communication overhead. Our results confirm improvements for important use cases compared to previous solutions. Furthermore, this approach can support co-scheduling tasks across multiple applications. - Open Access
- 5 June 2021
- Published by Springer Nature in International Journal of Parallel Programming
- Vol. 49 (6) , 867-885
Portable Node-Level Parallelism for the PGAS Model
The Partitioned Global Address Space (PGAS) programming model brings intuitive shared memory semantics to distributed memory systems. Even with an abstract and unifying virtual global address space it is, however, challenging to use the full potential of different systems. Without explicit support by the implementation node-local operations have to be optimized manually for each architecture. A goal of this work is to offer a user-friendly programming model that provides portable performance across systems. In this paper we present an approach to integrate node-level programming abstractions with the PGAS programming model. We describe the hierarchical data distribution with local patterns and our implementation, MEPHISTO, in C++ using two existing projects. The evaluation of MEPHISTO shows that our approach achieves portable performance while requiring only minimal changes to port it from a CPU-based system to a GPU-based one using a CUDA or HIP back-end.
- 5 June 2021
- Published by Springer Nature in International Journal of Parallel Programming
- Vol. 49 (6) , 867-885
Open AccessPortable Node-Level Parallelism for the PGAS Model
The Partitioned Global Address Space (PGAS) programming model brings intuitive shared memory semantics to distributed memory systems. Even with an abstract and unifying virtual global address space it is, however, challenging to use the full potential of different systems. Without explicit support by the implementation node-local operations have to be optimized manually for each architecture. A goal of this work is to offer a user-friendly programming model that provides portable performance across systems. In this paper we present an approach to integrate node-level programming abstractions with the PGAS programming model. We describe the hierarchical data distribution with local patterns and our implementation, MEPHISTO, in C++ using two existing projects. The evaluation of MEPHISTO shows that our approach achieves portable performance while requiring only minimal changes to port it from a CPU-based system to a GPU-based one using a CUDA or HIP back-end.
- Open Access
- 31 July 2020
- Published by Springer Nature
- p. 103-142
DASH: Distributed Data Structures and Parallel Algorithms in a Global Address Space
- 31 July 2020
- Published by Springer Nature
- p. 103-142
Open AccessDASH: Distributed Data Structures and Parallel Algorithms in a Global Address Space
- 1 March 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 353-357
Utilizing Heterogeneous Memory Hierarchies in the PGAS Model
Emerging technologies such as non-volatile or 3D-stacked memory significantly impact the design of future high performance computing systems. To keep up with the increasing core count, relying only on DRAM is inefficient due to its static power consumption. Modern HPC architectures feature a heterogeneous memory hierarchy with different capacities and capabilities. This paper addresses two challenges. First, we need programming models to abstract the complexity of the underlying heterogeneous memory hierarchy, while still giving explicit control to domain experts. We propose the concept of memory spaces to model a heterogeneous memory hierarchy and integrate it into a PGAS-like programming model. Second, we need to understand the impact of different memory capabilities on the performance of scientific applications. An experimental evaluation with a series of benchmarks, conducted on a Intel KNL platform, reveals that proper data placement on specific types of memory achieves significant speedup.
- 1 March 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 353-357
Utilizing Heterogeneous Memory Hierarchies in the PGAS Model
Emerging technologies such as non-volatile or 3D-stacked memory significantly impact the design of future high performance computing systems. To keep up with the increasing core count, relying only on DRAM is inefficient due to its static power consumption. Modern HPC architectures feature a heterogeneous memory hierarchy with different capacities and capabilities. This paper addresses two challenges. First, we need programming models to abstract the complexity of the underlying heterogeneous memory hierarchy, while still giving explicit control to domain experts. We propose the concept of memory spaces to model a heterogeneous memory hierarchy and integrate it into a PGAS-like programming model. Second, we need to understand the impact of different memory capabilities on the performance of scientific applications. An experimental evaluation with a series of benchmarks, conducted on a Intel KNL platform, reveals that proper data placement on specific types of memory achieves significant speedup.
- 1 March 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 18-25
A Portable Multidimensional Coarray for C++
Fortran Coarrays are a well known data structure in High Performance Computing (HPC) applications. There have been various attempts to port the concept to other programming languages that have a wider user base outside of scientific computing. While a popular implementation of the partitioned global address space (PGAS) model is Unified Parallel C (UPC), there is currently no portable implementation of Coarrays for C++. In this paper a portable version is presented, which is closely based on the Coarray C++ implementation of the Cray Compiling Environment. In this work we focus on a common subset of all proposed features by Cray. Our implementation utilizes the distributed data structures provided by the DASH library, demonstrating their universal applicability. Finally, a performance evaluation shows that our proposed Coarray abstraction adds negligible overhead and even outperforms native Coarray Fortran.
- 1 March 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 18-25
A Portable Multidimensional Coarray for C++
Fortran Coarrays are a well known data structure in High Performance Computing (HPC) applications. There have been various attempts to port the concept to other programming languages that have a wider user base outside of scientific computing. While a popular implementation of the partitioned global address space (PGAS) model is Unified Parallel C (UPC), there is currently no portable implementation of Coarrays for C++. In this paper a portable version is presented, which is closely based on the Coarray C++ implementation of the Cray Compiling Environment. In this work we focus on a common subset of all proposed features by Cray. Our implementation utilizes the distributed data structures provided by the DASH library, demonstrating their universal applicability. Finally, a performance evaluation shows that our proposed Coarray abstraction adds negligible overhead and even outperforms native Coarray Fortran.
- 15 September 2016
- Published by Springer Nature
- p. 361-377
Tool Support for Developing DASH Applications
- 15 September 2016
- Published by Springer Nature
- p. 361-377
Tool Support for Developing DASH Applications