D2 3 Computing e-Infrastructure cost calculations and business _models_vam1-final

More documents

Recommendations

Info

e-‐FISCAL: www.efiscal.eu EC Contract Number: 283449 Increasing complexity of data & difficulty of data gathering Full cost accounting • Detailed line item costs • Lengthy calculations • Actual costs, since purchase Total cost of ownership • Detailed line item cost projections over the lifetime of equipment /project • Predicted total cost over given timescale E-‐FISCAL annual cost • Estimated year operating costs • Annualized cost of simulated infrastructure • E-‐FISCAL yearly cost Past Today Future Figure 8: Comparing TCO, FCA and e-‐FISCAL methodology in cost calculations As it can be inferred by the graph above Full Cost Accounting may end up using obsolete assumptions, whereas TCO requires making assumptions on the whole lifecycle of the equipment (which in the case of machine rooms can be decades), bringing in their own uncertainties. Our proposed hybrid methodology is suitable for estimating the level and breakdown of current costs as well as allowing for the accurate projection into the short-‐term. The approach balances the level of detail and the accuracy of necessary costs against the effort required to gather and report the cost data. Moreover, it accounts for costs incurred but not paid by the site. As Intersect 360 comments in both reports of 2012 and 2013, several of their respondents stated that either the facilities cost were not in their budget or that they did not pay for every item (such as power or building/floor space) in their budget. 3.3 Research setting In order to be able to approximate the overall cost of the dedicated European HTC and HPC infrastructures we would ideally need cost information from the total population of dedicated European HTC and HPC infrastructures. However, as it is in practical terms very difficult to get data from every single HTC or HPC centre (i.e. to run a census), our analysis is expected to come up with robust and reliable results if there is an acceptable rate of participating entities. Having said that, it is not the absolute percentage of contributors that matter the most; their diversity is what it counts more. In order to be able to extrapolate the findings to the whole population of dedicated European HTC and HPC infrastructures, an adequate representation of sites/centres in different countries, of different size and of different nature (HTC/HPC) is needed. Getting input from sites/centres of different size would also permit the identification of any economies of scale effects. e-‐FISCAL : Financial Study for Sustainable Computing e-‐Infrastructures Deliverable D2.3 – Computing e-‐Infrastructures cost estimation and analysis – Pricing and Business models 36
e-‐FISCAL: www.efiscal.eu EC Contract Number: 283449 In order to cover as many cases (sites/centres) as possible, the method to collect data has been decided, as early as in the proposal phase, to be that of the questionnaire. The use of questionnaires permits the collection of information in a homogeneous and standard way and is less demanding in resources compared to interviews. Also, as the cost data asked is highly likely not in the possession of only one person but needs the cooperation of several people from different posts, the use of questionnaire facilitates fragmented provision of information as information is registered in a given document at any time convenient to the respondent. 3.4 Survey Instrument The developed questionnaire for the scope of the study was designed to provide the relevant inputs to satisfy requirements of the cost model presented in section 3.2.1. The questionnaire design needed to balance between conciseness and the ability to accurately distinguish important differences in cost structures of the centres. To strike an optimal balance, different versions of the questionnaire were tested internally in several iterations, after which the final version was sent to a small set of voluntary participants in the survey. The survey instrument went through an intensive review process that could be summarized as follows: • 1 st Draft questionnaire (already in August 2011) • Internal review and feedback from all e-‐FISCAL partners • 2 nd Draft questionnaire • Presentation of the questionnaire to the dedicated e-‐FISCAL workshop in EGI Technical Forum in Lyon (21/9/2011) 34 • Collection of both on-‐site and off-‐site comments from participants • Pre-‐final questionnaire (October 2011) • Pilot testing of the questionnaire (1 HTC site, 1 HPC centre and one mixed HTC/HPC centre • Final questionnaire integrating the comments from the above centres (November 2011) The review process performed concluded on a survey instrument that: • Balances the effort in gathering and reporting information with the usefulness of the information retrieved • Ensures that all necessary data for running the costing model is requested • Ensures that information available by other sources (e.g. EGI portal) is used to avoid asking respondents the same questions and to avoid duplication of work • Includes questions that are both clear and easily understood in a unilateral way from all respondents • Is applicable to both HTC and HPC centres 34 A copy of this version of the questionnaire is found in https://www.egi.eu/indico/conferenceTimeTable.py?confId=452#20110921. e-‐FISCAL : Financial Study for Sustainable Computing e-‐Infrastructures Deliverable D2.3 – Computing e-‐Infrastructures cost estimation and analysis – Pricing and Business models 37
Page 1 and 2: e-‐FISCAL: www.efiscal.eu EC
Page 35: e-‐FISCAL: www.efiscal.eu EC
Page 87 and 88:
e-‐FISCAL: www.efiscal.eu EC
Page 89 and 90:
Page 91 and 92:
Page 93 and 94:
Page 95 and 96:
Page 97 and 98:
Page 99 and 100:
Page 101:
show all

D2 3 Computing e-Infrastructure cost calculations and business _models_vam1-final

Create successful ePaper yourself

Delete template?

Save as template?