Project objectives

Green computing is currently a popular trend and need among European and world computing centers. The basic assumption is to use the existing computing infrastructure in such a way to minimize electricity consumption without significantly affecting the efficiency of calculations and the quality level of the services provided.

CI TASK, together with the ETI Faculty, actively participates in research and development (R&D) of modern technologies enabling the optimization of the use of electricity in large-scale calculations. The following project goals are currently being implemented:

 

  1. Methods development and software development for multi-criteria optimization (energy, performance) for High Performance Computing (HPC) for CPU and GPU computing units, using hardware power-cap mechanisms.
  2. Construction of electricity consumption models for the analysis of large-scale data sets (Big Data), using multi-node computer systems, with particular emphasis on modern data processing platforms (e.g. Apache Spark), using the Green Lab, created at CI TASK.

read more

A special laboratory was established to achieve the objectives of the Green Lab project. It consists of 16 computing nodes, located in a server racks, connected by an Ethernet 1Gb/s network. During computational experiments, electricity consumption is monitored on several levels: power strips (PDUs), power supplies and directly processors (CPUs), and memory (RAM). The figure below shows a diagram of the laboratory:

}

Completion dates

Internal project, leaded since 2017 with the participation of the ETI Department. Estimated time perspective: 5 years. The costs include the maintenance of the GreenLab laboratory (CI TASK) and expenses related to the publication of test results. Implementation from own funds.

 

i

List of scientific publications related to the project

[1] Paweł Czarnul, Jerzy Proficz, Krzysztof Drypczewski, “Survey of Methodologies, Approaches, and Challenges in Parallel Programming Using High-Performance Computing Systems”, Scientific Programming, page 1-19, Jan 2020.

[2] Paweł Czarnul, Jerzy Proficz, Adam Krzywaniak, „Energy-Aware High-Performance Computing: Survey of State-of-the-Art Tools, Techniques, and Environments”, Scientific Programming, volume 2019, page 1-19, Jan 2019.

[3] Adam Krzywaniak, Paweł Czarnul, Jerzy Proficz, „Extended investigation of performance-energy trade-offs under power capping in HPC environments”, page -, Jan 2019.

[4] Krzysztof Drypczewski, Jerzy Proficz, Andrzej Stepnowski, „Influence of YARN Schedulers on Power Consumption and Processing Time for Various Big Data Benchmarks”, TASK Quarterly, volume 22, page 313-319, Jan 2019.

[5] Adam Krzywaniak, Jerzy Proficz, Paweł Czarnul, „Analyzing energy/performance trade-offs with power capping for parallel applications on modern multi and many core processors”, Annals of Computer Science and Information Systems, volume 15, page 339-346, Jan 2018.

[6] Paweł Czarnul, Jarosław Kuchta, Paweł Rościszewski, Jerzy Proficz, „Modeling energy consumption of parallel applications”, Annals of Computer Science and Information Systems, volume 8, page 855-864, Jan 2016.

[7] Jerzy Proficz, Paweł Czarnul, „Performance and Power-Aware Modeling of MPI Applications for Cluster Computing”, Parallel Processing and Applied Mathematics, Part II, page 199-209, Jan 2016.