Parallel input/output (I/O) worklaod characterization studies are necessary to understanding the workload behaviors that dominate performance. This can translate into design principles and considerations that lead to high performance/cost implementations, by giving priority to the resource needs of these important factors. In this paper we present experimental results of an I/O workload charact...

This document describes the operating system support component in the Scalable I/O Initiative. Our e orts cover three critical areas of scalable, parallel I/O for for highperformance multicomputers: networking, memory management, and le and object store. We are leveraging o on-going research results in these three areas. Our main e orts in the scalable I/O initiative will be directed towards ha...

We give a quick presentation of the language CURIO, which gives a semantics to concurrent I/O in pure functional languages by way of modelling the API directly. Central to CURIO is the fact that an I/O model must obey a pre-condition – some broad axiomatic properties – so that program execution can be deterministic. We then present some mechanisms for modelling the dynamic allocation of data, c...

We present GPU-to-CPU callbacks, a new mechanism and abstraction for GPUs that offers them more independence in a hetero-ion for GPUs that offers them more independence in a heterogeneous computing environment. Specifically, we provide a method for GPUs to issue callback requests to the CPU. These requests serve as a tool for ease-of-use, future proofing of code, and new functionality. We class...

This document describes the operating system support component in the Scalable I/O Initiative. Our eeorts cover three critical areas of scalable, parallel I/O for high-performance multicomputers: networking, memory management, and le and object store. We are lever-aging oo ongoing research results in these three areas. Our main eeorts in the scalable I/O initiative will be directed towards hard...

SUMMARY Over the past several years, my colleagues and I have written a number of software libraries for fundamental computing tasks including I/O, memory allocation, container data types, and sorting. These libraries have proved to be good software building blocks and are used widely by programmers around the world. This success is due in part to a library architecture that employs two main in...

We discuss some techniques for preserving locality of reference in index spaces when mapped to memory units in a distributed memory architecture. In particular, we discuss the use of multidimensional address spaces instead of linearized address spaces, partitioning of irregular grids, and placement of partitions among nodes. We also discuss a set of communication primitives we have found very u...

identity and collective memory is one of the fascinating subjects in modern sociology. in iran, it arose again as modern and government-nation discourse emerged, and has been endlessly discussed since then. such subjects have assumed an increasing importance due to the multi-cultural and multi-ethical existence of iranian society and recent global developments. accordingly, the present study ha...