On the concept of simultaneous execution of multiple applications on hierarchically based cluster and the silicon operating system

Abstract

In this paper we present a novel cluster paradigm and silicon operating system. Our approach in developing the competent cluster design revolves around an execution model to aid the execution of multiple independent applications simultaneously on the cluster, leading to cost sharing across applications. The execution model should envisage simultaneous execution of multiple applications (running traces of multiple independent applications in the same node at an instant, without time sharing) and on all the partitions(nodes) of a single cluster, without sacrificing the performance of individual application, unlike in the current cluster models. Performance scalability is achieved as we increase the number of nodes, the problem size of the individual independent applications, due to non-dependency across applications and hence increase in the number of non-dependent operations( as the problem sizes of the applications get increased) and this leads to better utilization of the unused resources within the node. This execution model is very much dependent on the node architecture for performance scalability. This would be a major initiative towards achieving performance Cost-Effective Supercomputing.

Energetics based spike generation of a single neuron: simulation results and analysis

“…The parallel environment for the large scale simulation is facilitated by a hierarchy of host and node processes (Venkateswaran et al, 2008 ).…”

Towards modeling and integrated design automation of supercomputing clusters (MIDAS)

“…The design automation space for this research is taken to be the Memory-In-Processor SuperComputer-On-Chip (MIP SCOC) paradigm presented in [1,2], but can be extended to any general design space consisting of COTS and custom components. MIDAS developed cluster architecture is unconventional and takes major deviation as it is meant for simultaneous execution of multiple applications (from different users) and concurrent execution of traces of multiple applications in every node unlike the co-scheduling [6] and the IBM's SEQUOIA to be built by 2012 [7].…”

“…This is done by using the equivalent BEN-SIM graph of the application as it effectively captures the dynamics of the application. Although the MIDAS framework presented here and in [3][4][5] has a strong flavor of the MIP SCOC cluster architecture [1,2], it can be generalized to suit any architecture as the framework of both the model as well as the simulation is flexible. Figure 2 defines the dependency across various parameters in different design spaces of the supercomputing cluster.…”

“…The detailed host architecture can be referred from Fig. 8 in [1]. The architecture of the scheduler could be highly complex involving game theoretic optimizations or could be a simple first fit or best fit based scheduling strategy.…”