Algorithm-oriented design of efficient many-core architectures applied to dense matrix multiplication

Recent integrated circuit technologies have opened the possibility to design parallel architectures with hundreds of cores on a single chip. The design space of these parallel architectures is huge with many architectural options. Exploring the design space gets even more difficult if, beyond performance and area, we also consider extra metrics like performance and area efficiency, where the designer tries to design the architecture with the best performance per chip area and the best sustainable performance. In this paper we present an algorithm-oriented approach to design a many-core architecture. Instead of doing the design space exploration of the many core architecture based on the experimental execution results of a particular benchmark of algorithms, our approach is to make a formal analysis of the algorithms considering the main architectural aspects and to determine how each particular architectural aspect is related to the performance of the architecture when running an algorithm or set of algorithms. The architectural aspects considered include the number of cores, the local memory available in each core, the communication bandwidth between the many-core architecture and the external memory and the memory hierarchy. To exemplify the approach we did a theoretical analysis of a dense matrix multiplication algorithm and determined an equation that relates the number of execution cycles with the architectural parameters. Based on this equation a many-core architecture has been designed. The results obtained indicate that a 100 mm(2) integrated circuit design of the proposed architecture, using a 65 nm technology, is able to achieve 464 GFLOPs (double precision floating-point) for a memory bandwidth of 16 GB/s. This corresponds to a performance efficiency of 71 %. Considering a 45 nm technology, a 100 mm(2) chip attains 833 GFLOPs which corresponds to 84 % of peak performance These figures are better than those obtained by previous many-core architectures, except for the area efficiency which is limited by the lower memory bandwidth considered. The results achieved are also better than those of previous state-of-the-art many-cores architectures designed specifically to achieve high performance for matrix multiplication.

Influence of a lifestyle intervention in preschool children on physiological and psychological parameters (Ballabeina): study design of a cluster randomized controlled trial.

Childhood obesity and physical inactivity are increasing dramatically worldwide. Children of low socioeconomic status and/or children of migrant background are especially at risk. In general, the overall effectiveness of school-based programs on health-related outcomes has been disappointing. A special gap exists for younger children and in high risk groups. This paper describes the rationale, design, curriculum, and evaluation of a multicenter preschool randomized intervention study conducted in areas with a high migrant population in two out of 26 Swiss cantons. Twenty preschool classes in the German (canton St. Gallen) and another 20 in the French (canton Vaud) part of Switzerland were separately selected and randomized to an intervention and a control arm by the use of opaque envelopes. The multidisciplinary lifestyle intervention aimed to increase physical activity and sleep duration, to reinforce healthy nutrition and eating behaviour, and to reduce media use. According to the ecological model, it included children, their parents and the teachers. The regular teachers performed the majority of the intervention and were supported by a local health promoter. The intervention included physical activity lessons, adaptation of the built infrastructure; promotion of regional extracurricular physical activity; playful lessons about nutrition, media use and sleep, funny homework cards and information materials for teachers and parents. It lasted one school year. Baseline and post-intervention evaluations were performed in both arms. Primary outcome measures included BMI and aerobic fitness (20 m shuttle run test). Secondary outcomes included total (skinfolds, bioelectrical impedance) and central (waist circumference) body fat, motor abilities (obstacle course, static and dynamic balance), physical activity and sleep duration (accelerometry and questionnaires), nutritional behaviour and food intake, media use, quality of life and signs of hyperactivity (questionnaires), attention and spatial working memory ability (two validated tests). Researchers were blinded to group allocation. The purpose of this paper is to outline the design of a school-based multicenter cluster randomized, controlled trial aiming to reduce body mass index and to increase aerobic fitness in preschool children in culturally different parts of Switzerland with a high migrant population. Trial Registration: (clinicaltrials.gov) NCT00674544.

Chancen und Risiken der Arbeit im E-Business

E-Business, verstanden als ganzheitliche Strategie zur Reorganisation von Geschäftsprozessen, Strukturen und Beziehungen in Unternehmen, bietet für die Arbeitsgestaltung in einer digital vernetzten Welt Chancen und Risiken in Hinblick auf die Humankriterien. Empirische Untersuchungen in 14 Unternehmen zeigen „good practice“-Ansätze im B2B-Feld (Business-to-Business). Untersucht wurden die Tätigkeiten der elektronisch vernetzten Auftragsbearbeitung, des Web-, Content-Managements, der digitalen Druckvorlagenherstellung sowie der CAD- Bauplanzeichnung. Die beobachteten Arbeitsplätze zeigen, dass Arbeitsinhalte eher ganzheitlich und komplex gestaltet sind. Planende, ausführende, kontrollierende und organisierende Anteile weisen auf eine vielfältige Aufgabengestaltung hin, die hohe Anforderungen beinhaltet. Während alle beobachteten Tätigkeiten mit Aufnahme-, Erarbeitungs-, Verarbeitungs-, Übertragungs- und Weitergabeprozessen von Informationen zu tun haben, gibt es Differenzen in Bezug auf den Arbeitsumfang, den Zeitdruck, Fristsetzungen, erwartete Arbeitsleistungen sowie die Planbarkeit der Aufgaben. Die vorgefundenen Aufgabentypen (wenig bis sehr anforderungsreich im Sinne von Denk- und Planungsanforderungen) sind gekennzeichnet durch eine unterschiedlich ausgeprägte Aufgabenkomplexität. Interessant ist, dass, je anforderungsreicher die Aufgabengestaltung, je höher die Aufgabenkomplexität, je größer die Wissensintensität und je niedriger die Planbarkeit ist, desto größer sind die Freiräume in der Aufgabenausführung. Das heißt wiederum, dass bei zunehmenden E-Business-Anteilen mehr Gestaltungsspielräume zur Verfügung stehen. Die bestehenden Chancen auf eine humane Aufgabengestaltung sind umso größer, je höher die E-Business-Anteile in der Arbeit sind. Diese Wirkung findet sich auch bei einem Vergleich der Aufgabenbestandteile wieder. Die negativen Seiten des E-Business zeigen sich in den vorgefundenen Belastungen, die auf die Beschäftigten einwirken. Diskutiert wird die Verschiebung von körperlichen hin zu psychischen und vorrangig informatorischen Belastungen. Letztere stellen ein neues Belastungsfeld dar. Ressourcen, auf welche die Mitarbeiter zurückgreifen können, sind an allen Arbeitsplätzen vorhanden, allerdings unterschiedlich stark ausgeprägt. Personale, organisationale, soziale, aufgabenbezogene und informatorische Ressourcen, die den Beschäftigten zur Verfügung stehen, werden angesprochen. In Bezug auf die Organisationsgestaltung sind positive Ansätze in den untersuchten E-Business-Unternehmen zu beobachten. Der Großteil der untersuchten Betriebe hat neue Arbeitsorganisationskonzepte realisiert, wie die vorgefundenen kooperativen Organisationselemente zeigen. Die kooperativen Organisationsformen gehen allerdings nicht mit einer belastungsärmeren Gestaltung einher. Das vorgefundene breite Spektrum, von hierarchisch organisierten Strukturen bis hin zu prozess- und mitarbeiterorientierten Organisationsstrukturen, zeigt, dass Organisationsmodelle im E-Business gestaltbar sind. Neuen Anforderungen kann insofern gestaltend begegnet und somit die Gesundheit und das Wohlbefinden der Mitarbeiter positiv beeinflusst werden. Insgesamt betrachtet, zeigt E-Business ein ambivalentes Gesicht, das auf der Basis des MTO-Modells (Mensch-Technik-Organisation) von Uhlich (1994) diskutiert wird, indem vernetzte Arbeitsprozesse auf personeller, technischer sowie organisationaler Ebene betrachtet werden. E-business, seen as more than only the transformation of usual business processes into digital ones, furthermore as an instrument of reorganisation of processes and organisation structures within companies, offers chances for a human oriented work organisation. Empirical data of 14 case studies provide good practice approaches in the field of B2B (Business-to-Business). The observed work contents show, that tasks (e.g. order processing, web-, contentmanagement, first print manufacturing and architectural drawing) are well arranged. Executive, organising, controlling and coordinating parts constitute a diversified work content, which can be organised with high demands. Interesting is the result, that the more e-business-parts are within the work contents, on the one hand the higher are the demands of the type of work and on the other hand the larger is the influence on workmanship. The observed enterprises have realised new elements of work organisation, e.g. flexible working time, cooperative leadership or team work. The direct participation of the employees can be strengthened, in particular within the transformation process. Those companies in which the employees were early and well informed about the changes coming up with e-business work, the acceptance for new technique and new processes is higher than in companies which did not involve the person concerned. Structured in an ergonomic way, there were found bad patterns of behaviour concerning ergonomic aspects, because of missing knowledge regarding work-related ergonomic expertise by the employees. E-business indicates new aspects concerning requirements – new in the field of informational demands, as a result of poorly conceived technical balance in the researched SME. Broken systems cause interruptions, which increase the pressure of time all the more. Because of the inadequate usability of software-systems there appear in addition to the informational strains also elements of psychological stress. All in all, work contents and work conditions can be shaped and as a result the health and well-being of e-business-employees can be influenced: Tasks can be structured and organised in a healthfulness way, physiological strain and psychological stress are capable of being influenced, resources are existent and developable, a human work design within e-business structures is possible. The ambivalent face of e-business work is discussed on the basis of the MTO- (Mensch-Technik-Organisation) model (Ulich 1994). Thereby new and interesting results of researches are found out, concerning the personal/human side, the technical side and the organisational side of e-business work.

Research by design : a research strategy

The idea of an expressive component in research is important to the architectural industry. The expressive element - the possibility of expressing the qualitative aspects of the world and adding something new to the existing through experiments and proposals - is characteristic for the field. All research environments, in the science tradition and in the humanities, have their characteristics. On the one hand, they live up to certain common scientific and methodological criteria - originality and transparency – and on the other hand, they have different practices, using different methods. Research is ‘coloured’ by traditions and professions, and research in architecture should be coloured too, taking into consideration that the practice of architects stretches from natural science and sociology to art and that the most important way in which the architect achieves new cognition is through work with form and space – drawings, models and completed works. Probably all good design is informed by some kind of research – research- based design. But can research arise from design?