Magnus Jahre
Background and activities
Teaching
I am involved in teaching the following courses:
- TDT4255 Computer Design
- TDT01 Architecture of Computing Systems
I also supervise project and master thesis topics within computer architecture and design. Current project and master thesis topics are available at IDIs web pages. I often co-supervise projects and masters with local and national industry partners such as ARM, Nordic Semiconductor and Silicon Labs (formerly Energy Micro).
The topics and reports of my supervised master theses can be found on NTNU Open.
Research
The main goal of my research is to contribute to designing faster and more energy-efficient computers. More specifically, I investigate how computer hardware can specialize to the current application – to improve efficiency – while retaining sufficient generality to be efficient across diverse applications – to enable reuse. I am affiliated with the NTNU's Computer Architecture Lab (CAL) and currently serve as the deputy head of the Computing research group.
I am currently involved in the following research project:
- Project Manager and PI of the Balancing Compute and Memory Performance in Reconfigurable Accelerators with Analytical Modeling (BAMPAM) project. BAMPAM is a young research talents project funded by the Norwegian Research Council program IKTPLUSS.
We recently completed the TULIPP Horizon 2020 project. One of the key outcomes of TULIPP was the definition of a reference platform for high-performance low-power embedded image processing. Another key contribution is the STHEM utilities which makes it possible to non-intrusively estabilsh the energy consumption of source code constructs such as procedures and loops [video]. To make full use of STHEM, you need the Lynsyn power measurement unit which can be bought from Sundance.
I currently supervise/mentor the following PhD students and post docs:
- Lasse Agentoft Eggen (main supervisor)
- Fatemeh Ghasemi (main supervisor)
- Björn Gottschall (main supervisor)
- Joseph Rogers (main supervisor)
- Truls Asheim (co-supervisor)
- Even Låte (co-supervisor)
- Lahiru Rasnayake (co-supervisor)
I have supervised/metored the following PhD students and post docs:
- Main supervisor for Dr. Nico Reissmann (2012-2019), next employer NTNU IT
- Mentor for Dr. Asbjørn Djupdal (2013-2019), next employer NTNU
- Main supervisor for Dr. Yaman Umuroglu (2012-2018), next employer Xilinx
- Co-supervisor for Dr. Yahya Yassin (2012-2018), next employer Mode Sensors
- Mentor for Dr. Ananya Muddukrishna (2016-2018), next employer ÅF
- Mentor for Dr. Mohammed Sourori (2015-2017), next employer Accenture
- Co-supervisor for Dr. Odd Rune Strømmen Lykkebø (2012-2017), next employer Nnaisense
- Mentor for Post doc. Dr. Juan Manuel Cebrian (2012-2014), next employer UPC/BSC
- Mentor for Post doc. Dr. Nikita Nikitin (2013-2014), next employer Mentor Graphics
- Informal co-supervisor for Dr. Alexandru Ciprian Iordan (2008-2017), next employer ARM
My most recent research appears below. For more details regarding my publications, visit my Google Scholar profile or Cristin.
The following papers are in press:
- Lu Wang, Magnus Jahre, Almutaz Adileh, and Lieven Eeckhout. MDM: The GPU Memory Divergence Model. To appear in the 2020 International Symposium on Microarchitecture (MICRO).
- Xia Zhao, Magnus Jahre, and Lieven Eeckhout. Selective Replication in Memory-Side GPU Caches. To appear in the 2020 International Symposium on Microarchitecture (MICRO).
- Yahya Yassin, Magnus Jahre, Per Gunnar Kjeldsberg, Snorre Aunet, and Franky Catthoor. Fast and Accurate Edge Computing Energy Modeling and DVFS Implementation in GEM5 using System Call Emulation Mode. To appear in Journal of Signal Processing Systems.
- Magnus Jahre, Diana Göhringer and Philippe Millet (Editors). Towards Ubiquitous Low-power Image Processing Platforms. To be published by Springer.
Scientific, academic and artistic work
Displaying a selection of activities. See all publications in the database
2020
- (2020) HSM: A Hybrid Slowdown Model for Multitasking GPUs. ASPLOS'20: Proceedings of the Twenty-Fifth International Conference on Architectural Support for Programming Languages and Operating Systems.
2019
- (2019) DCMI: A Scalable Strategy for Accelerating Iterative Stencil Loops on FPGAs. ACM Transactions on Architecture and Code Optimization (TACO). vol. 16 (4).
- (2019) Modeling Emerging Memory-Divergent GPU Applications. IEEE computer architecture letters. vol. 18 (2).
2018
- (2018) GDP: Using Dataflow Properties to Accurately Estimate Interference-Free Performance at Runtime. IEEE Symposium on High-Performance Computer Architecture (HPCA). vol. 2018-February.
- (2018) Get Out of the Valley: Power-Efficient Address Mapping for GPUs. International Symposium on Computer Architecture.
2017
- (2017) FINN: A Framework for Fast, Scalable Binarized Neural Network Inference. Proceedings of the 2017 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays.
2016
- (2016) Efficient control flow restructuring for GPUs. International Conference on High Performance Computing & Simulation (HPCS).
- (2016) Random access schemes for efficient FPGA SpMV acceleration. Microprocessors and microsystems. vol. 47B.
2015
- (2015) Perfect Reconstructability of Control Flow from Demand Dependence Graphs. ACM Transactions on Architecture and Code Optimization (TACO). vol. 11 (4).
- (2015) ParVec: vectorizing the PARSEC benchmark suite. Computing. vol. 97 (11).
- (2015) Hybrid Breadth-First Search on a Single-Chip FPGA-CPU Heterogeneous Platform. 25th International Conference on Field Programmable Logic and Applications, FPL 2015, London, United Kingdom, September 2-4, 2015.
2014
- (2014) Optimized Hardware for Suboptimal Software: The Case for SIMD-aware Benchmarks. IEEE INTERNATIONAL SYMPOSIUM ON PERFORMANCE ANALYSIS OF SYSTEMS AND SOFTWARE.
- (2014) Graph-based Performance Accounting for Chip Multiprocessor Memory Systems. Proceedings of the 23rd International Conference on Parallel Architectures and Compilation Techniques (PACT).
- (2014) An Energy Efficient Column-Major Backend for FPGA SpMV Accelerators. 2014 32nd IEEE International Conference on Computer Design (ICCD).
2010
- (2010) Multi-level Hardware Prefetching Using Low Complexity Delta Correlating Prediction Tables with Partial Matching. Lecture Notes in Computer Science (LNCS). vol. 5 (1).
- (2010) DIEF: An Accurate Interference Feedback Mechanism for Chip Multiprocessor Memory Systems. Lecture Notes in Computer Science (LNCS).
2009
- (2009) A Quantitative Study of Memory System Interference in Chip Multiprocessor Architectures. 11th IEEE International Conference on High Performance Computing and Communications (HPCC 2009).
- (2009) A Light-Weight Fairness Mechanism for Chip Multiprocessor Memory Systems. Proceedings of the 6th ACM Conference on Computing Frontiers.