Add to ORKGDisk I/O transfer rates and overhead CPU times were measured as functions of buffer size and number of logically independent I/O channels for several operating systems and 16 I/O routines on the Cray-1 and CDC 7600 computers. By parameterizing the codes for a variable number of channels, buffer sizes, and words transmitted, the effect of these variables is observed for buffered, nonbuffered, and random-access I/O transmissions. To measure CPU-overlapped performance, I/O was performed concurrently with a pretimed compute loop. Rates, sector overhead, and CPU transmission speeds were calculated upon completion of I/O. Effects of memory blocking due to vector operations were observed. Methods and results are presented. 4 figures, 1 table
Although there are several extant studies of parallel scientific application request patterns, there...
This is an author's peer-reviewed final manuscript, as accepted by the publisher. The published arti...
This report compares the performance of different computer systems message passing. Latency and band...
257 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1993.There exists an increasing di...
Parallelization of I/O using multiple disk drives is warranted for achieving high performance on sys...
I/O is becoming the bottleneck in computer systems as the disparity between the speed of the I/O sys...
For a wide variety of computational tasks, disk I/O continues to be a serious obstacle to high perfo...
Recent studies have demonstrated that significant I/O operations are performed by a number of differ...
Input/output subsystem performance is currently receiving considerable research attention. Significa...
The CPUs, memory, interconnection network, operating system, runtime system, I/O subsystem, and appl...
The importance of a well tuned I/O subsystem is beyond doubt. In many cases, I/O may be the real bot...
We present a novel taxonomy that characterizes in a structured way the software and hardware tradeof...
Since the I/O subsystem is the bottleneck in external mergesort, I/O parallelism can result in subst...
Abstract: Rapid advances in processor performance have shifted the performance bottleneck to I/O sys...
The broadening disparity between the performance of I/O devices and the performance of processors an...
Although there are several extant studies of parallel scientific application request patterns, there...
This is an author's peer-reviewed final manuscript, as accepted by the publisher. The published arti...
This report compares the performance of different computer systems message passing. Latency and band...
257 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1993.There exists an increasing di...
Parallelization of I/O using multiple disk drives is warranted for achieving high performance on sys...
I/O is becoming the bottleneck in computer systems as the disparity between the speed of the I/O sys...
For a wide variety of computational tasks, disk I/O continues to be a serious obstacle to high perfo...
Recent studies have demonstrated that significant I/O operations are performed by a number of differ...
Input/output subsystem performance is currently receiving considerable research attention. Significa...
The CPUs, memory, interconnection network, operating system, runtime system, I/O subsystem, and appl...
The importance of a well tuned I/O subsystem is beyond doubt. In many cases, I/O may be the real bot...
We present a novel taxonomy that characterizes in a structured way the software and hardware tradeof...
Since the I/O subsystem is the bottleneck in external mergesort, I/O parallelism can result in subst...
Abstract: Rapid advances in processor performance have shifted the performance bottleneck to I/O sys...
The broadening disparity between the performance of I/O devices and the performance of processors an...
Although there are several extant studies of parallel scientific application request patterns, there...
This is an author's peer-reviewed final manuscript, as accepted by the publisher. The published arti...
This report compares the performance of different computer systems message passing. Latency and band...