Financial Study for Sustainable Computing e-Infrastructures

You are here

State of the Art

This table provides a list of papers relevant to the e-FISCAL work, which have been reviewed as part of the project state-of-the-art analysis. The papers mainly deal with financial aspects of High Throughput, High Performance, or Cloud Computing or other aspects related to the project such as Energy and Green IT.

If you are aware of a paper that is relevant and not listed below, please send us a note using the contact form. The project members are also active in the LinkedIn group established for discussions about ICT cost issues.

Paper

Summary

1. Do Clouds Compute? A Framework for Estimating the Value of Cloud Computing, Markus Klems et al, WEB 2008, LNBIP 22, pp. 110–123, 2009, Springer-Verlag Berlin Heidelberg 2009

In this paper, authors discuss the need for valuation of Cloud Computing, identify key components, and structure these components in a framework. Their proposed framework aims at assisting decision makers in estimating Cloud Computing costs and to compare these costs to conventional IT solutions.

2. Identification of a company’s suitability for the adoption of cloud computing and modelling its corresponding Return on Investment, Subhas Chandra Misra et al, Telecommunications Software Engineering: Emerging Methods, Models and Tools, Mathematical and Computer Modelling, Volume 53, Issues 3-4, February 2011, Pages 504-521, Elsevier Ltd.

The authors aim at helping companies analyze several characteristics of their own business as well as pre-existing IT resources to identify their favorability in the migration to the Cloud Architecture.  (Access only with subscription)

3. The Real Cost of a CPU Hour, Edward Walker, Computing Practices, April 2009, IEEE Computer Society

This paper proposes a modeling tool that can quantitatively compare the cost of leasing CPU time from online services to that of purchasing and using a server cluster of equivalent capability.

4. Cost-Benefit Analysis of High Performance Computing Infrastructures, Amril Nazir et al, 2010, IEEE2010

Authors carry out a total cost of ownership (TCO) analysis of a private resource system and a dedicated large HPC system. They compare the performance and cost-benefits of a dedicated HPC system in contrast to a private resource system. They conclude that a small private resource system with renting capability can provide a viable alternative to users to run their HPC applications, without the need to purchase and maintain a large, dedicated, HPC cluster system.

5. Cloud Computing for Research, Max Hammond et al, Curtis and Cartwright Ltd, June 2010

The scope of this report is to consider Cloud Computing for research in the areas of compute and storage. It covers the issues of Infrastructure as a Service (IaaS) and Platform as a Service but not Software as a Service. According to authors this report is intended to provide decision support as critical thinking is still required from researchers and institutions as to what data storage or compute solution is most appropriate given functional requirements, budget, security, reliability, trust, etc. as well as the cloud services currently on offer.

6. Green Data Center Program, Alan Crosswell, Presentation, Columbia University, June 2010

This presentation refers to the Green Data Center Program of Columbia University and discusses green best practices applied to the specific center. It contains detailed information and examples related to energy efficiency measurement.

7. Hyak Operating Costs and Comparison with Commercial Alternatives, University of Washington – e- Science Institute, Web article

In this article the cost of operating the Hyak shared compute facility at the University of Washington is explained. Moreover a comparison with some commercial alternatives is included.

8. What Does Grid Computing Cost?, Alek Opitz et al, Journal of Grid Computing, Feb 2008, 6:385–397, Springer Science & Business Media

In this article authors analyze the different general cost categories related to grid computing and determine the total costs of a resource provider. They give concrete numbers for the different cost categories and use these numbers to estimate the costs of the EGEE project and the Grid of the pharmaceutical company Novartis.

9. Worldwide electricity used in data centers, Jonathan G Koomey, Environmental Research Letters, IOP Publishing, September 2008

This study estimates historical electricity used by data centers worldwide and regionally on the basis of more detailed data than were available for previous assessments, including electricity used by servers, data center communications, and storage equipment.

10. The Economics Of The Cloud, Microsoft White paper, November 2010

This paper discusses the economics of scale implications for cloud computing as well as the concerns raised and cost implications for public and private clouds.

11. The Data Furnace: Heating Up with Cloud Computing, Jie Liu et al, June 2011

In this paper, authors argue that servers can be sent to homes and office buildings and used as a primary heat source. They call this approach the Data Furnace or DF and they claim that DFs create new opportunities for both lower cost and improved quality of service, if cloud computing applications can exploit the differences in the cost structure and resource profile between Data Furnaces and conventional data centers.

12. Cloud Computing and Grid Computing 360-Degree Compared, Ian Foster et al

 

This paper aims at comparing and contrasting Cloud Computing with Grid Computing from various angles. Moreover, it gives insights into the essential characteristics of both.

13. Why high-performance clouds are best kept in-house, By Rutrell Yasin, Mar 28, 2011

In this paper it is claimed that most commercial entities don’t have the infrastructure to handle the intensive workloads of high-performance computing and, if they do, it will probably be more expensive – in one case, 10 times more expensive – for them to run dedicated services than for some agencies to run their own private clouds.

14. Adoption Issues for Cloud Computing, Won Kim et al, 11th International Conference on Information Integration and Web-based Applications & Services  (iiWAS2009), Kuala Lumpur, Malaysia

In this paper the status of cloud computing today and various adoption issues, including costs, are discussed.

15. Computation Cost in Grid Computing Environments, Enis Afgan et al, 29th International Conference on Software Engineering Workshops (ICSEW'07), IEEE Computer Society

The goal of this paper is to develop a solid foundation in terms of viable options and questions one needs to consider when dealing with the variability found in pricing options within grid computing.

16. Cost-Benefit Analysis of Cloud Computing versus Desktop Grids, Derrick Kondo et al, 18th International Heterogeneity in Computing Workshop, May, 2009, Rome

In this paper, authors compare and contrast the performance and monetary cost-benefits of clouds for desktop grid applications, ranging in computational size and storage. They examine several research questions using performance measurements and monetary expenses of real desktop grids and the Amazon elastic compute cloud.

17. EGEE, an EGEE comparative study: Grids and Clouds evolution or revolution, Members of the EGEE-II Collaboration, 30/05/2008

This report compares grid and cloud computing services, taking a practical look at implementations of the Enabling Grids for E-sciencE (EGEE) project for grid and the Amazon Web Service (AWS) for cloud. Taking performance, scale, ease of use, costs, functionality and other aspects into consideration, the report looks at the overall opportunity that converging cloud and grid services can bring to users.

18. EGI-InSPIRE EU deliverable D2.6 - Integration of clouds and virtualisation into the European production infrastructure, Members of the EGI-InSPIRE Collaboration, 01/03/2011

This report aims to evaluate technologies such “Infrastructure as a Service’, ‘Platform as a Service’ and ‘Software as a Service’, understand how they relate to EGI, and build a foundation for the integration of cloud and virtualisation into the European production infrastructure. The report includes a cost analysis and comparisons to current market offers (chapter 5).

19. e- IRGSP2 EU deliverable D4.3 - Final legal issues report public, Members of the e-IRGSP2 Consortium, 28/01/2011

In this report (part b) the results of the financial exercise performed by the e-IRGSP2 is presented. The financial exercise aimed at estimating the cost of the EGI pan–European grid infrastructure for 2009 by extrapolating the cost of a few (seven) selected NGIs as example cases. Within the study the cost per CPU core hour under alternative CPU utilization rates is calculated and compared to the Amazon EC2 prices.

20. EGI-InSPIRE EU deliverable D2.7 - EGI sustainability plan, Members of the EGI-InSPIRE Collaboration, 31/03/2011

This report provides an updated taxonomy of services and a first outline of potential business models relevant to EGI serving. Within the business models identified sources of revenue and support are discussed such as fixed fee-based and usage-based models, free to use models, in-kind effort sources and public funding.

21. Growth in Data centre electricity use 2005 to 2010, Jonathan Koomey et al, CA: Analytics Press, August 1st 2011, Oakland

 

The study reports that the rapid rates of growth in data center electricity use that prevailed from 2000 to 2005 slowed significantly from 2005 to 2010, yielding total electricity use by data centers in 2010 of about 1.3% of all electricity use for the world, and 2% of all electricity use for the US.

22. Assessing trends over time in performance, costs, and energy use for servers. Koomey et al, CA: Analytics Press,  August 17th 2009, Oakland,

This article assesses trends in servers to help explain the driving forces affecting data center costs. It develops and documents detailed examples from available data, estimating costs and correcting them for inflation, and explaining the implications of the results.

23. Assessing trends in the electrical efficiency of computation over time, Koomey et al, CA: Analytics Press, Oakland August 17th 2009

This article explores the relationship between the processing power of computers and the electricity required to deliver that performance. More specifically, it estimates how many calculations historical and current computers were (or are) able to complete per kilowatt-hour of electricity consumed, which is one way to measure the electrical efficiency of computation over time.

24. Determining Total Cost of Ownership for Data Center and Network Room Infrastructure, White paper 6, Schneider Electric, Neil Rasmussen, 2011.

 

The paper presents a method for measuring total cost of ownership (TCO) of data center network room physical infrastructure and relates these costs to the overall information technology infrastructure with examples. Also, the cost drivers of TCO are quantified.

25. Cost Analysis of Current Grids and Its Implications for Future Grid Markets Grid Economics and Business Models, Marcel Risch and Jörn Altmann, Lecture Notes in Computer Science, Volume 5206/2008.

This paper analyzes the question whether using the Grid is financially advantageous, using the Amazon.com EC2 service as a reference. To perform this analysis, the costs of computing resources in different usage scenarios are calculated, if Grid resources and in-house resources are used. The comparison of the costs reveals that while the Grid is cheaper in the short term, it is not a good investment in the long term and, thus, the existence of a Grid economy will not lead to an end of ownership but rather to a reduction in in-house resources and more efficient resource usage.

26. Economic aspects of hybrid cloud infrastructure: User organization perspective, Oleksiy Mazhelis and Pasi Tyrväinen, 2011

In this paper, an analytical model of hybrid cloud costs is introduced, wherein the costs of computing and data communication are taken into account. Using this model, a cost-efficient division of the computing capacity between the private and the public portion of a hybrid cloud can be identified.

27. A Cost-Benefit Analysis of a Campus Computing Grid 2,Smith, Preston M. M.S., Purdue University, May 2011

In this thesis, the author presents a model for calculating the base cost for a core-hour of computation in Purdue University’s campus grid.  With a cost model, the author then analyzes the benefits gained from using the grid, based on the number of hours of delivered, number of computations completed, and the number of users and faculty members served.

28. Cost-effective HPC: The community or the cloud. IEEE Conference on Cloud Computing Technology and Science, Carlyle A. G. et al, December 2010

.

This paper is a case study of costs incurred by faculty end-users of Purdue University's HPC “community cluster” program. The authors develop and present a per node-hour cloud computing equivalent cost that is based upon actual usage patterns of the community cluster participants and is suitable for direct comparison to hourly costs charged by one commercial cloud computing provider. They find that the majority of community cluster participants incur substantially lower out-of-pocket costs in this community cluster program than in purchasing cloud computing HPC products.

29. eScience Institute, Cloud Economics: Visualizing AWS Prices over Time, Posted by billhowe | November 28, 2010

 

In this article, the authors focus on the steady decrease in prices offered by Amazon for their services. They comment that over time, the price to rent 1 unit of resources for three years of continuous usage has fallen dramatically as Amazon offered new instance types, offered new long-term pricing plans, and lowered prices outright across the board.

30. Magellan Final Report, December 2011 Magellan is a project funded through the U.S. Department of Energy (DOE) Office of Advanced Scientific Computing Research (ASCR). Its goal is to investigate the potential role of cloud computing in addressing the computing needs for the DOE Office of Science (SC). In the conclusions of the project’s final report it is stated, among others, that “the cost analysis shows that DOE centers are cost competitive, typically 3-7x less expensive when compared to commercial cloud providers”.
31. Recovery Oriented Computing (ROC): Motivation,
Definition, Techniques, and Case Studies,Paterson et al,
Computer Science Technical
Report UCB//CSD-02-1175, U.C. Berkeley, March 15, 2002.
 
In this paper authors deal with Recovery Oriented Computing (ROC). They claim that since a large portion of system administration is dealing with failures, ROC may reduce total cost of ownership. They argue that one to two orders of magnitude reduction in cost mean that the purchase price of hardware and software is now a small part of the total cost of ownership.
32. IMEX TCO Analysis High Peformance Computing. In this report IMEX Research presents the results of the evaluation of the three-year life cycle TCO (Total Cost of Ownership including Acquisition and Operation Costs) of a 512 node cluster in HPC environment using APPRO HyperBlade Servers based on AMD Opteron processors and 1U servers from Dell, Hewlett Packard, IBM and Sun.
33. Which is less expensive: Amazon or self-hosted? Charlie Oppenheimer, Matrix Partners Feb. 11, 2012 In this article the author discusses the economics of the choice between self-hosted and cloud provider (Amazon Web Servises – AWS) as a function of demand distribution.
34. “Cost analysis of cloud computing for research”, Hawtin et al, Final Report to EPSRC and JISC, 22/2/2012, Curtis+Cartwright Consulting Ltd. In this report researchers analyse the common view that cloud computing is about to save money. They suggest that this is certainly not clear for research computing. They conclude that on a pure price comparison, the more powerful cloud computing instances, rented on an hourly basis, appear to be one-and-a-half to two times more expensive per core-hour than well-managed, locally-provided clusters in modern data centres operating at high utilisation levels. However, other purchasing models (such as ‘Reserved Instances’) can reduce the costs to parity or better.
35. "Cloud computing as an innovation: Perception, attitude, and adoption" Lin A. et al. International Journal of Information Management (2012), http://dx.doi.org/10.1016/j.ijinfomgt.2012.04.001 This study aims to investigate how cloud computing is understood by IT professionals and the concerns that IT professionals have in regard to the adoption of cloud services. The study was carried out in Taiwan and used a survey by interview approach to understand IT professionals’ understandings and concerns about cloud computing. The findings of the study suggest that while the benefits of cloud computing such as its computational power and ability to help companies save costs are often mentioned in the literature, the primary concerns that IT managers and software engineers have are compatibility of the cloud with companies’ policy, IS development environment, and business needs; and relative advantages of adopting cloud solutions. The findings also suggest that most IT companies in Taiwan will not adopt cloud computing until the uncertainties associated with cloud computing, e.g. security and standardisation are reduced and successful business models have emerged.
36. “Cloud computing — The business perspective” Marston S et al Decision Support Systems 51 (2011), 176–189 In this article, authors identify the strengths, weaknesses, opportunities and threats for the cloud computing industry. They also identify the various issues that will affect the different stakeholders of cloud computing. They issue a set of recommendations for the practitioners who will provide and manage this technology. For IS researchers, they outline the different areas of research that need attention so to advice the industry in the years to come. Finally, they outline some of the key issues facing governmental agencies who, due to the unique nature of the technology, will have to become intimately involved in the regulation of cloud computing.
37. "Performance and Cost Assessment of Cloud Services” Paul Brebner and Anna Liu, Lecture Notes in Computer Science, , Volume 6568, Service-Oriented Computing, Pages 39-50, 2011 Authors have developed a Service Oriented Performance Modeling technology for modeling the performance and scalability of Service Oriented applications architected for a variety of platforms. Using a suite of cloud testing applications we conducted in-depth empirical evaluations of a variety of real cloud infrastructures, including Google App Engine, Amazon EC2, and Microsoft Azure. The insights from these experimental evaluations, and other public/published data, were combined with the modeling technology to predict the resource requirements in terms of cost, application performance, and limitations of a realistic application for different deployment scenarios.
38.  "Towards A European Roadmap for Datacentres and their Economic and Environmental Impact", Cordis (2011), Workshop Reportheld in Brussels in November 2011  This report discusses potential actions that could contribute towards improving efficiency in data centres and reducing costs. The report presents a discussion about ways that Europewould be mobilized to minimize both the environmental and economic impact of such centres while addressing the issue that the amount of data shows a considerable increasing tread. The strategies are discussed under two perspectives: The “low hanging fruit actions” and the long term programs. The report touches up the whole resource efficiency consideration of data centres (energy consumption, thermal dissipation, use of water for cooling, carbon footprint, the construction of data centers per se). The report considers critical the measuring of impact through metrics welcoming PUE as a first step, while acknowledging differences among data centres. Actions that could contribute towards decreasing power consumption include among others, applications software efficiency, hardware/software co-design, advances in processor design, new approaches in minimizing cooling requirements, improvements to uninterruptible power supplies (UPSs) and usage of renewable energy sources.
39. “Benchmarking the Amazon Elastic Compute Cloud (EC2)", Coffey et al, Major Qualifying Project Report submitted to the Faculty of the WORCESTER POLYTECHNIC INSTITUTE in partial fulfillment of the requirements for the Degree of Bachelor of Science , 9/3/2011 This project sought to determine whether Amazon EC2 is an economically viable environment for scientific research-oriented activities within a university setting. The methodology involved benchmarking the performance of the cloud servers against the best systems available at the Electrical and Computer Engineering (ECE) Department of WORCESTER POLYTECHNIC INSTITUTE (WPI). Results indicate that the newest ECE server outperformed the best EC2 instance by approximately 25% in most cases. A comprehensive cost analysis suggests that EC2 instances can achieve up to 60% better cost to performance ratios in the short-term when compared against the ECE servers. However, a long-term projected cost analysis determined that the overall cost of owning a large set of reserved instances comes out to almost 60% more than the cost of comparable in-house servers.
40. “On the Economics of Huge Requirements of the Mass Storage: a Case Study of the Agata Project”, Mendez Munoz et al, CLOSER 2011 - International Conference on Cloud Computing and Services Science, pp. 506-511.
This case study focuses in costs analysis of huge size mass storage under three options: a dedicated storage, a Grid storage and a Cloud storage service. It concludes that the costs savings depend more in the particularity of the problems than in general estimations. Moreover, the results show a lower total costs for the Grid option.
41. “Disk and Tape Storage Cost Models”, R. J. Moore et al, San Diego Supercomputer Center, University of California San Diego; La Jolla, CA, USA, 2007
This paper describes current estimates of both disk and tape storage based on operational experience at the San Diego Supercomputer Center which operates a large-scale storage infrastructure. Moreover, authors present their views on projected cost trends in both disk and tape and provide a comparison to current web-based commercial storage services.
42. Espida handbook “Expressing project costs and benefitsin a systematic way for investment in information and IT” University of Glasqow and JISC, 2007
The espida project aims at proposing a framework within which a good project proposal is prepared that permits that a clear communication between a proposer and a funder is achieved and the project proposal includes all costs and benefits clearly defined.
43. “Fair Benhcmarking for Cloud Computing Services”, L. B. Gillam et al, University of Surrey, March 2012. This project aims to at the practical characterisation of IaaS resources through means of assessing performance of three public Cloud providers and one private Cloud system, in relation to a set of established, reasonably informative, and relatively straightforward to run, benchmarks.
44. "Evolving the EGI Business Model" Sergio Andreozzi et al, EGI Deliverable D2.18, 25/05/2012 EGI’s main  contribution to  the digital European Research Area  is to deliver a world-class einfrastructure built as an open ecosystem that offers the opportunity for different actors to provide their own uniquely valuable  tools and services for the benefits of researchers. In order  for the whole EGI ecosystem to provide value sustainably, each actor needs to properly identify the most appropriate business model  for it  to operate with. This  report addresses  this need by  bringing together the information produced from a number of activities over the last year and providing a framework for discussing and generating business models for the actors within the EGI ecosystem. Concrete proposals and plans for the next two years are also provided."
45. "The Strategy of the Commons: Modelling 
for European Research" Matti Heikkurinen et al, eChallenges e-2012 Conference Proceedings, 17/10/2012
This paper presents a hybrid cost assessment method (e-FISCAL method) addressing the limitations of the standard methods (i.e. Total Cost of Ownership and Full Cost Accounting) and presents some of the initial results of the e-FISCAL study as an illustration of the kind of cost assessment issues high-utilisation rate ICT services should consider when choosing between different infrastructure options.
46. VENUS-C EU project deliverable, D3.10 - Future Sustainability Strategies, Members of the VENUS-C Consortium, chapter 4, 25/6/2012 This document identifies and discusses relevant aspects that influence the future sustainability of e-Science Cloud infrastructures. Moreover, the document presents a current market analysis along with an analysis of the VENUS-C cloud application community in terms of behaviour and requirements. The document presents three business models for the involvement of commercial cloud providers for e-Science purposes: a hybrid model where users interact with a local community cloud and an external commercial provider,  a similar one with the addition of a cloud broker in between, and one where users interact directly with an external commercial provider.
47. FedSM EU proejct deliverable D3.1: Business Models for federated e-Infrastructures, Owen Appleton, FedSM Consortium, This document presents a way to categorise different arrangements of federations in the eInfrastructure domain and looks at the business models that these arrangement imply. This includes developing some simple tools to assist in this process. The business models include invisible coordination, advisor, matchmaker, one stop shop and integrator roles.
48. Financial Aspects of Cloud Computing Business Models, Jaakko Jäätmaa, Aalto University School of Economics

The purpose of the study was to explore financial aspects of cloud computing business models from information technology (IT) services provider’s perspective. The financial aspects were divided into revenue model and related pricing mechanisms and cost structure and related cost accounting mechanisms according to business model ontology. Findings of the study suggested that each cloud service should have a distinct business model. Pricing of services changes with cloud computing and pay per use and subscription-based pricing mechanisms are most typical for cloud services. The pricing should be based on customer’s perceived value instead of production costs of services. A generic cloud computing
pricing mechanism that combines pay per use and  subscription mechanisms was created to better balance risk sharing between services provider and customer.

49. Computenext Cloud Service Marketplace Compunexts builds the next generation of compute power through the launch and operation of a federated cloud marketplace, leveraging multi-cloud connectors, to enable consumers and providers of cloud services to participate in the search, purchase, and utilization of the most efficient cloud resources available
50. Gartner Special Report on Cloud Computing - Cloud Services Brokerage, This report presents cloud services brokerage (CSB). CSB plays an intermediary role in cloud computing and makes it easier for organizations to consume and maintain cloud services, particularly when they span multiple providers. This research defines CSBs and outlines the capabilities that are needed in this role.