The CloudLightning Conference presentation recordings are now available on the CloudLightning YouTube channel.
Prof. John Morrison (UCC) – “Self-Organisation As A Cloud Resource Management Strategy”
Cloud Resource Management is becoming increasingly challenging with the advent of hyperscale computing and the proliferation of heterogeneous hardware. Meanwhile, resource utilisation continues to remain low resulting in high energy consumption per executed instruction. This talk by Prof. John Morrison suggests a self-organised approach to resource management in an attempt to successfully address these challenges.
Prof. Dieter Kranzlmueller (LMU Munich) – “Is Heterogeneity a solution to the HPC dilemma?”
High performance computing is an important and useful tool for many scientific domains, and many users demand highest performance for their applications. Today, most HPC machines provide a heterogeneous architecture, mixing different processors, accelerators, and other components specific for the expected load of the customers and following the latest trends in computer architectures. The HPC machines at the Leibniz Supercomputing Centre (LRZ), such as currently SuperMUC Phase 1 and Phase 2, are providing general purpose CPUs only, allowing a wide variety of users to utilize them for their computing tasks. With the replacement of SuperMUC currently being planned, the question is what to use in the future. Prof. Dieter Kranzlmueller provides an insight into the current and predicted state of supercomputing, and the consequences for the users.
Brian Boyle (HEANet) – “The Past, Present and (Possible) Future of Ireland’s National Education and Research Network”
HEAnet is Ireland’s National Education and Research Network (NREN), providing innovative services to institutions involved in education in Ireland. Advancements in network infrastructure in recent years have allowed NRENs like HEAnet to provide services over wide area networks which were previously only practical through direct local networks. As we all come to rely on networked (or cloud) based resources for day to day activities, this education network infrastructure has become vital for many aspects of campus life, with HEAnet’s services meshed very closely with campus services. This presentation by Brian Boyle seeks to explore the long term significance of this collective move towards a cloud based e-Infrastructural future.
Alison Kennedy (Hartree Centre) – “Supercomputing and Cloud – Competitors Or Collaborators?”
The Hartree Centre is a supercomputing centre, focused on bringing the benefits of high performance computing, high performance data analytics and cognitive computing to industry, to deliver “innovation at scale” and competitive advantage through digital transformation. The centre has recently purchased one of the largest supercomputers in the world on which to undertake their work and offer services. As an industrially-focused supercomputing centre, how does Hartree sees itself working with cloud providers and cloud-based solutions, to support UK industry?
Dr. Ravi Manumachu (UCD) – “Data-Parallel Applications On Multicore And Manycore Platforms – New Challenges and Novel Solutions”
Performance and energy efficiency are now critical concerns in high performance scientific computing. To address these twin concerns while continuing to provide unprecedented computational performance, HPC systems today have tight integration of energy-efficient multicore CPU processors and accelerators (GPUs, Intel Xeon Phis, FPGAs, etc). However, this tight integration has created formidable challenges for model and algorithm developers. In this talk, Dr. Ravi Reddy Manumachu focusses on the new complexities introduced in modern homogeneous parallel platforms composed of multicore and manycore processors such as resource contention and non-uniform memory access (NUMA). He presents latest advances that address the challenges posed by the complexities and discusses bi-objective optimization of data-parallel applications on such platforms for performance and energy.
David Power (vScaler) – “OpenStack For Scientific Research”
OpenStack is the leading open source IaaS platform, powering many of the world’s most notable science and research organisations. In this presentation David Power explores OpenStack as a cloud platform for scientific workloads and talks about the challenges, pitfalls and potential opportunities to deliver flexible infrastructure for research computing.
Dr. Christos Papadopoulos-Filelis (DUTH) – “HPC In The Cloud: Scientific Computations And Simulation”
A large number of HPC applications migrate to Cloud Environments. These HPC applications include various Engineering and Physics type simulations. Gravitational N-body simulation is one of the most challenging and computationally intensive applications used to study evolutionary phenomena of galaxies or cosmological structures. The requirements of this type of simulation are introduced and discussed with respect to emerging service delivery models of Cloud environments, such as: WaaS and CloudLightning. Moreover, discussions about other types of Engineering and Physics simulations that can be performed in the Cloud along with advantages and disadvantages are given.
Dr. Anthony Ventresque (UCD) – “Multi-Objective Virtual Machine Reassignment For Large Scale Data Centres”
Optimising the data centres of large IT organisations is a challenging problem for classical optimisation techniques, especially given the size of data centres. It is even more complex as (i) enterprise data centres are composed of various hosting departments with their own preferences and (ii) reassignment solutions can be evaluated from various independent dimensions. In this talk Dr. Anthony Ventresque shows how difficult this problem is for classical algorithms and how to address it using hybrid techniques.
Dr. Konstantinos Giannoutakis (CERTH) – “Simulation of Heterogeneous Cloud Infrastructures”
During the last years, except from the traditional CPU based hardware servers, hardware accelerators are widely used in various HPC application areas. More specifically, Graphics Processing Units (GPUs), Many Integrated Cores (MICs) and Field-Programmable Gate Arrays (FPGAs) have shown a great potential in HPC and have been widely mobilised in supercomputing and in HPC-Clouds. In this presentation Dr. Konstantinos Giannoutakis focuses on the development of a cloud simulation framework that supports hardware accelerators. The design and implementation of the framework are also discussed.