Parallel Distributed Infrastructure for Minimization of Energy

News

ParaDIME shortlisted for the 2016 Innovation Radar Prize!

Wed, 2016-07-20

Our innovation "power-estimator for heterogeneous CPU architectures" has been short-listed for the Innovation Radar Prize 2016 of the Horizon2020 Program! This puts us among the top 10 Research and Innovation Actions of the ICT subprogram for innovation potential.

At the end of July, the Commission will publish the profiles of the short-listed projects on their Innovation Radar website where you can vote for us. The top 4 projects will be then invited to the ICT Proposers Day on 26 September in Bratislava for presentation in front of a jury. Keep a close eye on the Innovation Radar website, and don't forget to vote! 

ParaDIME project on Science Node

Wed, 2016-04-13

 

An article about the ParaDIME project has been published on Science Node website. The Science Node is a free online publication, jointly funded by organizations in the US and Europe.

 

Please, find the article by following this link A new ParaDIME for energy-efficient computing or by downloading the PDF file here below.


ParaDIME’s innovations offer significant energy savings for data centres

Wed, 2015-10-28

Researchers from the ParaDIME research project, coordinated by Barcelona Supercomputing Center, have successfully developed a number of methodologies which enable savings in data centre energy consumption ranging from 30% to 60%. These methodologies tackle several critical power-related research challenges, from using different current and future devices to inter-datacentre VM (virtual machine) scheduling.

At the programming model level, the focus has been on shifting from the shared memory model to an actor-based, message-passing programming model which enables programmers to achieve greater energy efficiency and become more energy aware. There are two illustrative approaches:

•        Tailored programming solutions for heterogeneous GPU/CPU architectures. By directly managing the GPU through optimized code, energy consumption can be reduced by roughly 80%. However, this requires in-depth specialist knowledge, which reduces programmability. ParaDIME has developed techniques based on domain-specific languages (DSL) which generate code for both the CPU and GPU, resulting in energy savings of up to 40% while, crucially, empowering a far greater number of programmers to utilize these innovative architectures.

•        Tools for power and cost awareness that estimate the power requirements of a single process being run in a virtualized environment. These tools can also be used for user-based pricing models, energy-aware task scheduling and as an indicator for how many heterogeneous resources are consumed by an application.

At the runtime level, ParaDIME has developed a large, decentralized infrastructure of small data centres that provide heating and hot water. This is motivated by the efficiency gains demonstrated by the project’s industrial partner Cloud&Heat. ParaDIME researchers have developed:

·         The multi-datacentre scheduler: this schedules jobs across different data centres, striking a balance between data centre workloads and heating/cooling necessities, resulting in the reduction of CO2 emissions and energy consumption of up to 50%.

·         An intra-datacentre scheduler: technologies have been developed to reduce the time needed to reactivate virtual machines and their migration costs. Parts of this work are under review by the QEMU – an open-source machine emulator and virtualizer – community. Institutions using QEMU to virtualize their workload will be able to benefit from ParaDIME-optimized virtual machine migration code. Furthermore, ParaDIME has contributed a feature to track changes to block devices that has already been incorporated into the latest Linux kernel.

At the hardware level, ParaDIME researchers have proposed and simulated several methodologies for improving energy-efficiency of the future computing node, including:

·         Scheduling of tasks to heterogeneous cores (e.g. big.LITTLE processors, or systems that combine FPGA, GPU and CPU cores). On average, a 40% saving in energy consumption can be achieved by combining FPGA, GPU and CPU cores as opposed to a multicore processor. ParaDIME scheduling also reduces 20% of the power and 35% of the energy on average across different types of heterogeneous platforms. ParaDIME has also researched power estimation tools for a variety of core types. 

·         Aggressively lowering the supply voltage. Energy is saved by combining this with low-overhead error detection and correction techniques. In addition, ParaDIME researchers have explored this methodology for circuits built with future devices. The ParaDIME methodology saves up to 60% of the energy consumed by the L1 data cache.

Diagram showing ParaDIME infrastructure

                                                                          Figure 1 ParaDIME infrastructure          

About the ParaDIME project

ParaDIME ("Parallel Distributed Infrastructure for Minimization of Energy") was a three-year research project launched in September 2012 with a total budget of €3.2M, including €2.5M funding from the European Commission's Seventh Framework Programme. The project was coordinated by Barcelona Supercomputing Center (BSC) and partners were IMEC (Belgium), Technische Universität Dresden (Germany), Université de Neuchâtel (Switzerland) and Cloud&Heat (Germany).

The objective of this European project was to attack the power-wall problem by radical software-hardware techniques that are driven by future circuit and device characteristics on the hardware side, and by a programming model based on message-passing, and in a smart scheduling of the workload of data centres on the software side.

For further information, visit www.paradime-project.eu

ParaDIME @ HiPEAC CSW, 2015

Wed, 2015-09-23

ParaDIME was invited to present its results related to Power and Energy estimation at the System-Level at the TACLe Thematic Session. Link to the talk: https://www.hipeac.net/events/activities/7288/tacle/#fndtn-program. This talk was well attended by 55 people from 34 different institutions.

 

Abstract of the talk:

This talk was about the work done in the ParaDIME FP7 project about how to design and implement energy-aware computing and power estimation at the system-level for various types of low-power processors prototypes and to help improving energy efficiency of future computing systems. Power and energy consumption of processors are steadily increasing and the work performed by them at the data centers is not proportional to the power dissipated, where every µA is a revenue for the entity. On the one hand, the hardware community is proposing various methodologies to address this issue such as low-power processors, heterogeneity, etc. to reduce the power of the servers. On the other hand, the system software community proposes mechanisms such as virtual machines (VMs), work-load scheduling, etc. to increase the utilization of the processor. To properly evaluate the impact of these mechanisms, we need a fast and accurate power estimation tool at the system-level.

 

The link of the video will updated soon. 

ParaDIME @ ARM Research, Cambridge

Mon, 2015-09-21

ParaDIME disseminated its research results as a part of the invited talk series at the ARM Research in Cambridge on 18th of September.  This talk was attended by people from both ARM Cambridge and and from ARM, Austin, USA through a live feed.

First part of the talk was focused on how to design and implement energy-aware computing and power estimation in servers, MPSoCs and various types of data centers to help achieve an energy efficient computing future. Power and energy consumption of data centers are steadily increasing and the work performed by the data centers is not proportional to the power dissipated, where every µA is a revenue for the entity. On the one hand, the hardware community is proposing various methodologies to address this issue such as low-power processors, heterogeneity, etc. to reduce the power of the servers. On the other hand, the software community proposes mechanisms such as virtual machines (VMs), work-load scheduling, etc. to increase the utilization of the processor. In order to properly evaluate the impact of these mechanisms, we need an accurate power monitoring and estimation tool at the hardware host level, the VM level and the system-level. Achieving the target exascale performance and designing a green cloud computing infrastructure require the design of dynamic and smart techniques that recognize the hardware-software characteristics and optimization of the trade-off among performance, energy, and power in an application-aware manner. Second part of the talk was focused on a novel heterogeneous platform with the integration of three devices (CPU, GPU, FPGA) into a single system, i.e. Trigeneous platforms, to efficiently accelerate and to minimize energy of computation intensive applications in both high-performance computing and embedded system domains.

ParaDIME’s research paper has been nominated for Best Paper Award at IEEE-SoCC 2015 Conference

Tue, 2015-09-22

The paper entitled as FAcET: Fast and Accurate Power/Energy Estimation Tool for CPU-GPU Platforms at Architectural Level authored by Santhosh Kumar Rethinagiri, Oscar Palomar, Javier Arias Moreno, Osman Unsal and Adrian Cristal has been nominated the Best Paper Award at IEEE-SoCC 2015. This paper proposes a powerful power estimation tool at the system-level and a necessary tool for both application and architecture designers.

 

Abstract the paper:

This paper proposes a novel fast and accurate architectural-level tool to estimate power and energy (FAcET) for heterogeneous (CPU-GPU) system architecture based platforms. FAcET consists of two components. The first is a set of generic parametrizable power models generated by characterizing the functional-level activities for different blocks of the chosen platforms. The second is a simulation-based architectural-level prototype that uses SystemC (JIT) simulators to accurately evaluate the parameters of the corresponding power models of the first component. The combination of the two components leads to a novel power and energy estimation methodology at the architectural level that provides a better balance between speed and accuracy. The efficacy of the FAcET tool is verified against measurements taken on real board platforms, which consist of low-power ARM quad-core processors (Cortex-A7, -A9 and -A15), NVIDIA GPUs (Quadro 1000M, Quadro FX5600, Tegra K1, and GTX480) and heterogeneous platforms (NVIDIA Tegra3 and NVIDIA Jetson TK1). Power and energy estimation results obtained with FAcET deviate in less than 3.6% for quad-core processors, 6.5% for GPU, 10% for heterogeneous multiprocessor based systems from the measurements and estimation is 15x faster than state-of-the-art tools.

 

About the conference:

In its 27 years of history, the IEEE International System-on-Chip Conference (SOCC) has been the premier forum for sharing advances in system-on-chip (SoC) technologies, designs, tools, test, verification and applications. Held at changing locations in the USA, Europe and Asia, SOCC is attracting researchers and engineers from all over the world to exchange knowledge, share experiences and establish collaborations with colleagues. Each year, the IEEE-SoCC Programme Committee selects 3 papers to be nominated for the best paper award across 300˖ papers.

ParaDIME on the Workshop on the Future Energy in ICT Research Agenda

Mon, 2015-09-21
The Workshop on the Future Energy in ICT Research Agenda brought together in Bristol on the 15th of September academic and industrial leaders to focus on the challenge of energy consumption in ICT. The main workshop event featured presentations on current challenges and research in low energy of ICT from fundamental physics to HPC, and provided networking opportunities to discuss where attention should be focused.
 
ParaDIME partner Dr. Jens Struckmeier from Cloud&Heat presented the work of setting up a decentralized green data center for cloud computing. Introducing the concept and challenges of reusing the computational heat for heat purposes in houses and apartment complexes. Dr. Struckmeier also pointed out why decentralized datacenter allow for shorter response times and enable "real time" cloud applications required for future technologies (industry 4.0, autonomous driving, virtual reality, and the tactile internet).
 

 
After the talk there was a vital question and answer session showing the interest in the technology. The data of the other speakers (i.e. Prof. Paul Douglas, University of Glasgow and Boris Grot, University of Edinburgh) allow the conclusion that despite all the very successful efforts in reducing energy consumption of ICT the overall energy consumption of ICT will continue to grow because of the extreme increase of volume of data and data processing outgrowing the efficiency efforts. Therefore the concept of reusing the energy is complementary to all energy saving efforts and not contradictory.
 
Cloud&Heat provides a distributed Infrastructure-as-a-Service cloud based on servers that are located in eco-friendly residential or commercial buildings, powered by OpenStack. The set-up is straightforward: servers, in self-contained fireproof cabinets, are installed in the basements of private and commercial buildings. The cloud-heaters are connected via broadband internet fibre cable connections. To allay concerns about housing the servers in this new way, data is decentralized, triple-replicated and encrypted. The company offers single-rack installments with normal broadband connections and larger installations with connections to multiple backbones for redundancy. Between one and six cabinets (“data safe”) form an individual OpenStack deployment and they all share the same authentication through one OpenStack identity (code-named Keystone) service.
 
The business model rests on a foundation of energy efficiency, competitive pricing and data protection that meets strict German and European standards. The model is two-pronged: one set of customers have a cost-effective cloud and the other save on cost-efficient heating. Cutting costs on the cooling of servers while providing heat for homes is proving energy efficiency that could never be achieved with classical centralized data center. Our solution enables a PUE of 1.06 plus the energy is reused for heating houses which further reduces the CO2 footprint. We developed a concept to match server workload with heating demand inventing the concept of briques as computational and heating units.
 
Cloud&Heat (former AoTerra) has been awarded for its innovations i.e. 2013 with the Saxon Environmental Award, was finalist in the German Industry Innovation Award and won the German data center price (Deutscher Rechenzentrumspreis) for energy efficient data center in 2015.
 
In this talk Cloud&Heat gave a brief overview over their solution of a hybrid water and air cooled microdatacenter for cloud applications. And also will introduced the attendees to challenges and solutions of matching cloud demand to the heating and warm water demand of the buildings.

A new paper presented in PPPJ 2015

Mon, 2015-09-14
http://www.paradime-project.eu/system/files/news/images/pppj15.jpg
Sebastian Ertel presented the paper "Ohua: Implicit Dataflow Programming for Concurrent Systems", authored by Sebastian Ertel, Christof Fetzer and Pascal Felber, at the PPPJ 2015, held from 8 to 10 September in Melbourne (Florida, USA).
 
2015 International Conference on Principles and Practices of Programming on the Java Platform: virtual machines, languages, and tools (PPPJ’15) – the 12th conference in the PPPJ series – provides a forum for researchers, practitioners, and educators to present and discuss novel results on all aspects of programming on the Java platform including virtual machines, languages, tools, methods, frameworks, libraries, case studies, and experience reports.

ParaDIME team at PhD retreat Goslar

Fri, 2015-09-04
This year's PhD retreat was organized by the Dresden University of Technology (TUD), the University of Neuchâtel (UniNE), as well as the Braunschweig University of Technology. PhD students as well as post-docs had a chance to share their findings in fields such as SGX, Fault Tolerance, Parallel and Distributed Computing, Energy Efficiency and BFT.
 
ParaDIME members gave talks and discussed about various challenges faced in the ParaDIME project.
 

 
After attending the talks, participants had an opportunity to participate in the marshmallow challenge. In eighteen minutes, teams had to build the tallest free-standing structure out of 20 sticks of spaghetti, one yard of tape, one yard of string, and one marshmallow. The marshmallow needed to be on top. Out of 6 teams, only one team managed to build a standing structure. 
 
During one of the evenings, participants were also invited to walk to the local "Steinberg Alm zum Rösner" restaurant, where they enjoyed the regional German food. Overall, the retreat was a great success as it provided a wonderful opportunity to present research results as well as research plans to fellow PhD students.

ParaDIME presents research on hardware support for efficient message passing at ICPP-2015

Wed, 2015-09-02
http://www.paradime-project.eu/system/files/news/images/img_3136-cropped.jpg

This month, a ParaDIME paper was presented at the 44th International Conference on Parallel Processing (ICPP), held from 1-4 of September 2015, in Beijing, China. 

The paper entitled DiMP: Architectural Support for Direct Message Passing on Shared Memory Multi-cores” was authored by Ruben Titos-Gil, Oscar Palomar, Osman Unsal and Adrian Cristal, from Barcelona Supercomputing Center. It proposes and evaluates a set of ISA extensions, and their associated hardware support, aimed at reducing some of the overheads that message passing workloads suffer when executed on a commodity shared-memory multicore.

ParaDIME at the ICT Energy Summer School

Mon, 2015-07-20
The 6th edition of the NiPS Summer School took place in Fiuggi (Italy) from 7 to 12 July. This year's edition was entitled "ICT-Energy: Energy consumption in future ICT devices" and it was aimed at teaching the bases of the science of efficient ICT through 4 thematic groups of lectures:
  • Basic on the physics of energy transformations at micro and nanoscales
  • Introduction to energy harvesting and distributed autonomous mobile devices
  • Software and energy aware computing
  • High performance computing and systems

ParaDIME researcher Santhosh Rethinagiri presented two out of four High performance computing systems sessions: "Introduction to data-centers" and "Tools and methodologies for energy-aware data-centers".  

The presentations can be found in the website of the Summer School. Please, visit the ICT-Energy website, Facebook page and Twitter timeline to find out more information. A Facebook group from the Summer School was also created.






 

ParaDIME project is member of the Consortium of the coordination activity ICT-Energy.

Participating in the live Twitter chat "Less energy consumption in ICT"

Mon, 2015-06-22

ICT-Energy, a coordinated activity where ParaDIME project belongs, hosted a live Twitter chat on the 18th of June in the framework of the Sustainable Energy Week and the Micro-Energy Day. The Twitter chat could be followed by the hashtag #LessEnergyICT. Below these line you can read the Storify of the live chat.

 

 

 

 

During the week, ParaDIME partner Barcelona Supercomputing Center offered information to the visitors to the MareNostrum supercomputer about micro-energy. Micro-energy refers to the energy that is often disregarded as unimportant but actually plays a significant role in our daily life. As an example… when you run out of battery in your mobile phone and really need to make that call!!! The amount of energy involved in this case is really very small compared to the energy required to drive a car, but you definitely notice when it's not there…

 

ParaDIME shows off latest research about dynamic message processing in the Actor Model at DAIS

Thu, 2015-06-18

This month, a ParaDIME paper was presented at the 10th International Federated Conference on Distributed Computing Techniques (DAIS), held from 2-5 of June in Grenoble, France. The DAIS conference, part of the IFIP Distributed Computing Techniques (DisCoTec) series of federated conferences, is one of the leading international venues to discuss all aspects of distributed applications and systems.

The paper entitled “Dynamic Message Processing and Transactional Memory in the Actor Model” was authored by Yaroslav Hayduk, Anita Sobe and Pascal Felber from the University of Neuchâtel. It discusses strategies for processing messages concurrently in the Actor Model in cases of high contention.

ParaDIME presents a paper about memory management at ISCA Symposium

Mon, 2015-06-15
This week, one ParaDIME paper about memory management will be presented in the 42nd International Symposium on Computer Architecture (ISCA). ISCA is the premier forum for new ideas and experimental results in computer architecture and this year is held in Portland, Oregon from 13 to 17 of June.
 
The paper is titled “Redundant Memory Mappings for Fast Access to Large Memories” and it has been written by Vasileios Karakostas (from BSC/UPC) together with researchers of the University of Wisconsin-Madison, Microsoft Research and BSC. This paper suggests a hardware/software co-design that leverages ranges of pages to reduce the overhead of virtual memory.
 
Please, feel free to check out our Publications section to find this and other papers and scientific publications.

10 minutes with... Malte Schneegass, Cloud&Heat

Wed, 2015-06-10
Malte Schneegass received his engineering degree in computer and automation technology  from the University of Applied Science, Dresden (Germany). In 2012 he joined Cloud&Heat Technologies, a provider of cloud-based computing services, with the waste heat produced by the servers also being used to heat buildings and water. In May 2015, Cloud&Heat received the award Deutschen Rechenzentrumspreis 2015 in the category “Newly built energy-efficient and resource-efficient data centers”.
 
1. Could you explain us why the Cloud&Heat has a different approach to the cloud computing services?
Cloud&Heat is a distributed data center which combines the businesses of a cloud services provider and providing heating devices for residential and industrial buildings. The cloud servers are installed directly in the properties to be heated. Multiple smaller data centers are connected via the internet to form a virtual data center. Cloud users can take profit from this decentralized infrastructure by programming fail over mechanisms from one location to another and build up reliable and highly available services. 
 
Nowadays, our efforts are dedicated to manage and maintain the distributed data center. We are focused on developing the product, from the server components to the heating system, as well as the connectivity and scalability.
 
On the other hand, we are really happy with the feedback received from the heat market, which has demonstrated strong interest in our product.
 
2. How can the scientific research led to products or how can it reach the industry?
It’s not easy to reach the industry with research results. Usually there is still a long way to go to make them economically valuable for businesses. In our case, for example, we still need to continue our research and devote a lot of efforts to make a product that meets the industrial needs.  We are constantly improving our product and are able to establish a new innovative product in the market.
 
3. What led you to work in this field? What do you like about it?
I studied automation of systems and processes. My current work is about automating a highly distributed heating system that uses cloud computing as fuel. Before implementation, conceptional work with models and simulation was necessary. Afterwards, we shaped the plan with a lot of interdisciplinary work, computer science and HVAC engineering/control loops. It is thrilling to see how the different technologies create the synergy effect. I like to work on something that makes our daily life smarter and more sustainable at the same time. 
 
4. What are the key technical challenges which need to be tackled in order to achieve more energy-efficient computing systems?
With the installation of distributed data centers we have done a great step forward. The next step is to make it as energy-efficient as possible: this is something we already do in the ParaDime project by implementing and improving smart-scheduling algorithms in our data centres. However, the computer architecture and server architecture based on electronics loose heat. Either we are able to re-use this energy if not it is wasted heat that could be used for free warm water to shower or for heating purposes.