Cloud Resource Management is becoming increasingly challenging with the advent of hyperscale computing and the proliferation of heterogeneous hardware. Meanwhile, resource utilization continues to remain low resulting in high energy consumption per executed instruction. This talk will suggest a self-organized approach to resource management in an attempt to successfully address these challenges.
Achieving high resource utilization in internet-scale datacenters is key to cloud providers, while also of high significance to scientific discovery in supercomputing, where we are increasingly witnessing scientific simulations co-existing with data analytics and visualization jobs. In this talk, I will walk you through two pilot systems prototyped by IBM Research – Ireland, embodying novel computer architectures enabling – among others – new and unprecedented levels of resource utilization.
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 we will show how difficult this problem is for classical algorithms and how to address it using hybrid techniques.
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. This presentation focuses on the development of a cloud simulation framework that supports hardware accelerators. The design and implementation of the framework are also discussed.