Automatic Data Allocation and Buffer Management for Multi-GPU Machines

Multi-GPU machines are being increasingly used in high-performance computing. Each GPU in such a machine has its own memory and does not share the address space either with the host CPU or other GPUs. Hence, applications utilizing multiple GPUs have to manually allocate and manage data on each GPU. Existing works that propose to automate data allocations for GPUs have limitations and inefficiencies in terms of allocation sizes, exploiting reuse, transfer costs, and scalability. We propose a scalable and fully automatic data allocation and buffer management scheme for affine loop nests on multi-GPU machines. We call it the Bounding-Box-based Memory Manager (BBMM). BBMM can perform at runtime, during standard set operations like union, intersection, and difference, finding subset and superset relations on hyperrectangular regions of array data (bounding boxes). It uses these operations along with some compiler assistance to identify, allocate, and manage data required by applications in terms of disjoint bounding boxes. This allows it to (1) allocate exactly or nearly as much data as is required by computations running on each GPU, (2) efficiently track buffer allocations and hence maximize data reuse across tiles and minimize data transfer overhead, and (3) and as a result, maximize utilization of the combined memory on multi-GPU machines. BBMM can work with any choice of parallelizing transformations, computation placement, and scheduling schemes, whether static or dynamic. Experiments run on a four-GPU machine with various scientific programs showed that BBMM reduces data allocations on each GPU by up to 75% compared to current allocation schemes, yields performance of at least 88% of manually written code, and allows excellent weak scaling.

Research Data Management — Activities and Challenges Report on the Knowledge Exchange Workshop

Workshop Research Data Management – Activities and Challenges 14-15 November 2011, Bonn The Knowledge Exchange initiative organised a workshop to highlight current activities and challenges with respect to research data management in the Knowledge Exchange partner countries and beyond. The workshop brought together experts from data centres, libraries, computational centres, funding organisations, publishing services and other institutions in the field of research and higher education who are working to improve research data management and encourage effective reuse of research data. A considerable part of the programme was dedicated to sharing perspectives from these communities, leading to the development of a roadmap of practical actions for the Knowledge Exchange initiative, partner organisations and other stakeholders to progress over the next two years. On the first day, principal investigators and project managers from a great variety of recent projects shared their insights on objectives and methods for improving data management ranging from discipline-specific to more general approaches. A series of short presentations of selected projects was followed by an extensive poster session that functioned as a “trade fair” of current trends and activities in the field of research data management. Moreover, the poster session offered ample network opportunities for participants. The second day was dedicated to intensive group discussions looking at a number of data management challenges. First the most important findings from the "Surfboard for 'Riding the Wave'" report were presented. This included the state of the art on activities and challenges in the field of research data management. The subgroups will concentrate on the following key themes: funding, incentives, training and technical infrastructure. These discussions culminated in the identification of practical recommendations for future cooperation on practical as well as on strategic levels that should be taken forward by the KE partner organisations and beyond. These activities aim to improve the sustainability of services and infrastructures at both national and international levels.

Status of hilsa (Tenualosa ilisha) management in the Bay of Bengal: an assessment of population risk and data gaps for more effective regional management

Assessment of country status papers on hilsa fisheries presented at the BOBP – IGO Chittagong, Bangladesh 2010. Assessment of status hilsa management in Bangladesh, India and Myanmar. Brief recommendations of potential follow-up activities that could enhance management. Risk assessment of hilsa in each country with Productivity Susceptibility Analysis (PSA). Summary of new approach to assess ecological risk.

Research data management and openness: the role of data sharing in developing institutional policies and practices

Purpose: To investigate the relationship between research data management (RDM) and data sharing in the formulation of RDM policies and development of practices in higher education institutions (HEIs). Design/methodology/approach: Two strands of work were undertaken sequentially: firstly, content analysis of 37 RDM policies from UK HEIs; secondly, two detailed case studies of institutions with different approaches to RDM based on semi-structured interviews with staff involved in the development of RDM policy and services. The data are interpreted using insights from Actor Network Theory. Findings: RDM policy formation and service development has created a complex set of networks within and beyond institutions involving different professional groups with widely varying priorities shaping activities. Data sharing is considered an important activity in the policies and services of HEIs studied, but its prominence can in most cases be attributed to the positions adopted by large research funders. Research limitations/implications: The case studies, as research based on qualitative data, cannot be assumed to be universally applicable but do illustrate a variety of issues and challenges experienced more generally, particularly in the UK. Practical implications: The research may help to inform development of policy and practice in RDM in HEIs and funder organisations. Originality/value: This paper makes an early contribution to the RDM literature on the specific topic of the relationship between RDM policy and services, and openness – a topic which to date has received limited attention.

Data management and data analysis in the large European projects GEHA (GEnetics of Healthy Aging) and NU-AGE (NUtrition and AGEing): a bioinformatic approach

The aging process is characterized by the progressive fitness decline experienced at all the levels of physiological organization, from single molecules up to the whole organism. Studies confirmed inflammaging, a chronic low-level inflammation, as a deeply intertwined partner of the aging process, which may provide the “common soil” upon which age-related diseases develop and flourish. Thus, albeit inflammation per se represents a physiological process, it can rapidly become detrimental if it goes out of control causing an excess of local and systemic inflammatory response, a striking risk factor for the elderly population. Developing interventions to counteract the establishment of this state is thus a top priority. Diet, among other factors, represents a good candidate to regulate inflammation. Building on top of this consideration, the EU project NU-AGE is now trying to assess if a Mediterranean diet, fortified for the elderly population needs, may help in modulating inflammaging. To do so, NU-AGE enrolled a total of 1250 subjects, half of which followed a 1-year long diet, and characterized them by mean of the most advanced –omics and non –omics analyses. The aim of this thesis was the development of a solid data management pipeline able to efficiently cope with the results of these assays, which are now flowing inside a centralized database, ready to be used to test the most disparate scientific hypotheses. At the same time, the work hereby described encompasses the data analysis of the GEHA project, which was focused on identifying the genetic determinants of longevity, with a particular focus on developing and applying a method for detecting epistatic interactions in human mtDNA. Eventually, in an effort to propel the adoption of NGS technologies in everyday pipeline, we developed a NGS variant calling pipeline devoted to solve all the sequencing-related issues of the mtDNA.

Data Management Policies and Issues with Life Science Data

Data management and sharing are relatively new concepts in the health and life sciences fields. This presentation will cover some basic policies as well as the impediments to data sharing unique to health and life sciences data.