Resource-Aware Computing - Summer 15
- Term
- Summer 15
- Lecturer
- Univ.-Prof. Dr. Michael Bader, Dr. Josef Weidendorfer, Emily Mo-Hellenbrand, M.Sc., Alexander Pöppl, M.Sc., Isaías Alberto Comprés Ureña
- Time and Place
- : Initial Meeting: January 21st, 16:00, Room: MI 02.09.023
- Kickoff: t.b.a.
- Presentations: See Schedule section
- Audience
- Students from Master Informatics (IN2107), Computational Science and Engineering (IN2183), and Bachelor Informatics (IN0014)
- Tutorials
- -
- Exam
- -
- Semesterwochenstunden / ECTS Credits
- 2 SWS (2S) / 4 Credits
- TUMonline
- Master Seminar Resource-Aware Computing (IN2183,IN2107,IN0014)
News
- Max. number of participants: 12
Description
In the last ten years the period of vast increases in processing power mostly achieved by increasing the clock frequency of a processor has come to an end. Instead, computer architectures are getting more complex in order to accommodate the growing demand for processing power. Modern CPUs typically have a wide range of SIMD instructions for fine-grained data parallelism, and are capable of executing several threads on each of their several cores. Memory accesses are passed through multiple cache levels to hide memory access latencies. In addition to that, hardware specialized in performing massively parallel computations is getting more and more popular. Examples are GPUs and accelerators such as the Xeon Phi. In the HPC context, several nodes, each with its own CPU(s) and GPU(s) may be joined into a cluster.
Regular programming techniques and paradigms are no longer sufficient to fully utilize this hardware. Frameworks such as OpenCL take the structure and heterogeneity of the underlying hardware into account and provide the programming environment to expose all available resources, such as GPUs and accelerators. Novel approaches (such as invasive computing) expose the allocation of resources to the users, allowing them to request required resources and, by offering the reallocation of resources at runtime, enabling them to adapt to changing computing demands.
The behavior of the hardware at runtime also needs to be considered. Modern Cluster architectures are not necessarily capable to run at peak utilization 100% of the time. To avoid the overheating of the hardware and the resulting degradation of the silicon, the clock frequency of the CPU may be drastically reduced, or single nodes may even be shut down completely for a time. Taking these problems into account is an additional challenge developers face today.
Background: The seminar will (partly) discuss research executed in the collaorative research unit Invasive Computing funded by the German Research Foundation (DFG).
Schedule
| Date | Description | Material |
| 21.01.2015 | Initial Meeting | - |
| 1st week | Kick-Off Meeting (actual time t.b.a.) | - |
| 27.04.2015 | Submission of paper outline | - |
| 3rd week | Flash-Talk Meeting (actual time t.b.a.) | - |
| 7th week | Submission of 1st draft to supervisor (voluntary) | - |
| 05.06.2015 | Submission of paper for review (Release Candidate) | - |
| 12.06.2015 | Submission of paper reviews | - |
| 10th week | Presentation meeting 1 (actual time t.b.a.) | - |
| 10th week | Presentation meeting 2 (actual time t.b.a.) | - |
| 11th week | Presentation meeting 3 (actual time t.b.a.) | - |
| 11th week | Presentation meeting 4 (actual time t.b.a.) | - |
| 12th week | Presentation meeting 5 (actual time t.b.a.) | - |
| 12th week | Presentation meeting 6 (actual time t.b.a.) | - |
| 03.07.2015 | Submission of final paper | - |
Topics (preliminary)
- Invasive Computing: Concepts
- Invasive Computing: The Framework
- Dark silicon (hardware, current Trends, etc)
- Faults and Fault-Tolerance
- Resource management for shared-/distributed-memory architectures
- Resource management for heterogeneous architectures
- Invasive algorithms
- Invasive Network-on-a-Chip architectures (iNoCs)
- Hardware-aware security mechanisms
- Heterogeneous computing frameworks: OpenCL
- Characterization of problems and extraction of patterns
- Power Saving by Dynamic Voltage and Frequency Scaling (DVFS) and Power Capping (PC)
- Resource aware runtime systems (for shared-/distributed-memory architectures)
Note: Additional topics are possible, the current list is not final.
Literature
Consider the Invasive Computing homepage. We encourage independent research and review of the available literature.
Requirements
- Independent literature research
- Review of two papers written by fellow students
- 30 minute presentation with 15 minute discussion session
- Paper with 5-7 pages (max 9 pages) in total. IEEE-Format (see link below) is required.
- Mandatory attendance to all presentations. In case attendance is impossible due to good reasons, a prior notification to the supervisors is necessary. Doctor's attest needed in case of sickness.
Deadlines
- April 27th, Monday: submission of paper topic and outline
- June 5th, 23:59, Friday: submission of paper for review (Release Candidate)
- June 12th, 23:59, Friday: submission of paper reviews
- Jul 3rd, 23:59, Friday: submission of final paper
Note: Failure to meet these deadlines may result in grade deduction or failure of the course.