P. Pierce, "The NX/2 Operating System", Proceedings of the Third Conference on Hypercube Concurrent Computers and Applications, 1988, pp. 384-390.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....pre copying in the V kernel [10] copy on reference in Accent [14] and flushing dirty pages to disk in Sprite [5] While our basic technology for remote task creation is similar to these systems, none of them addressed the problem of creating multiple tasks or processes at the same time. The NX 2 [8] operating system supported concurrent remote process creation. The address spaces are created by loader in a spanning tree fashion. Performance is function of , and it is even further improved by using advanced communication mechanisms. Resulting performance ranged from as high as a fraction of a ....

Pierce, P., "The NX/2 Operating System", Proceedings of the Third Conference on Hypercube Concurrent Computers and Applications, vol 3, pp 384-390, January 1988.

....is required to map threads to PEs. With local queues, each PE is exactly like a uniprocessor. Indeed, this is used in the Unix compatible OSF 1 AD system [634] Other example systems that use this approach include distributed memory systems like PEACE on SUPRENUM [505, 506] NX 2 on the iPSC 2 [460], SIMPLEX on the nCUBE [335] the MOOSE experimental system [497] the original Cosmic Cube [513] and others [146] The Sylvan system went so far as to add a dedicated kernel processor to each node [91] one of the responsibilities of this processor was to perform local scheduling of the node s ....

....balancing by diffusion 4.1.2 [254] nCUBE partitioning with buddy system, local queues 3.2.1 [258] NETRA hierarchical structure for balanced mapping 4.1.1 [119] NQS batch queueing on partitioned systems 7.1.1, 3.2.3 [323] NX 2 (iPSC 2) partitioning into subcubes, local queues 3.2.1, 4. 1 [460] Orion (on M 3 ) family scheduling with pools of processors 4.2 [90] OSF 1 AD local Unix like scheduling plus load balancing 4.1, 4.1.2 [634] Paragon from Intel hierarchical partitioning with optional gang scheduling or space slicing 7.2.1 [288] PASM from Purdue partitioned multistage ....

P. Pierce, "The NX/2 operating system". In 3rd Conf. Hypercubes, Concurrent Comput., & Appl., vol. I, pp. 384--390, Jan 1988.

....use of the most attractive features of a number of existing message passing systems, rather than selecting one of them and adopting it as the standard. Thus, MPI1 has been strongly influenced by work at the IBM T. J. Watson Research Center by Bala, Kipnis, Snir and colleagues [1,2] Intel s NX 2 [17], Express [16] nCUBE s Vertex [14] and PARMACS [11,12] Other important contributions have come from Zipcode [18,19] Chimp [6,7] PVM [8,20] and PICL [9] One of the objectives of this paper is to promote a discussion within the concurrent computing research community of the issues that must ....

Paul Pierce, "The NX/2 Operating System," Proceedings of the Third Conference on Hypercube Concurrent Computers and Applications, pp. 384--390 (ACM Press, 1988).

....calls are supported at the host. These features were fixed in NX2 which provides full symmetry between host and node operations. 18 3.4. The Intel iPSC2 Computer 3.4.1. Intel iPSC2 Programming: NX2 NX2 is the current Intel operating system for the iPSC 860 and Paragon computers [4,5], see Table 5. The NX2 operating system introduced some small changes to the NX1 system, and added extra capabilities such as interrupt driven communication. Again NX2 represents each processor by an integer node number in the range [0, P 1] but now the host is denoted by the integer P . As in ....

P. Pierce, "The NX/2 Operating System", Proceedings of the Third Conference on Hypercube Concurrent Computers and Applications, January 1988.

....are several other interesting observations which can be made. Consider first the write behavior as shown in Figures 2(c) and 2(d) 1. The data transfer rate for the 1 MB file is approximately 1240 KB sec. The maximum point to point communication transfer rate of the iPSC 2 is specified by Intel [PIER88] to be 2800 KB sec and has been measured by others [BOMA89] at 2660 KB sec. Thus this output rate is slightly less than 50 of the basic message passing communication rate. 2. The data transfer rate for the 8 MB file is 750 800 KB sec or 610 KB sec depending on where the file header is relative ....

P. Pierce, "The NX/2 Operating System", Proc. 3rd Conf. on Hypercube Concurrent Computers and Applications, Pasadena, Jan. 1988, 384-390.

....include thousands of processors. On such multicomputer architectures, each node runs an instance of the operating system kernel. These differents instances communicate via message passing. Currently available multicomputers are generally featured with very simple operating system kernels (ex: NX [Pier88] on iPSC 2, Helios [Garn87] on transputers) offering mostly communications primitives. Services such as user interaction and device access are provided via separate host computers. The multicomputer is generally used as a high performance co processor connected to workstations and mainly for ....

Paul Pierce, "The NX/2 Operating System," in , Intel Scientific Computers, Portland, OR, (88/06), pp. 51-57. CS/EX-88-417

.... The information in this document for the iPSC 860 architecture and the NX 2 operating system was taken from example programs [1] manual pages [2] and manuals [3] made available by the Technical Publications Department of the Intel Supercomputer Systems Division as well as miscellaneous papers [4] [5] Information for the AP1000 was taken from the available programming manuals [6] 7] 8] 9] 10] 11] 12] 13] produced by Fujitsu Research Labs 1 . 1.3 Nomenclature and document format This document will use the nomenclature cell to refer to each individual processing unit of the ....

Paul Pierce, "The NX/2 Operating System," Proceedings of the Third Conference on Hypercube Concurrent Computers and Applications 1, 384--390.

....Nonetheless, good performance in this approach relies on an GID match being the common case, making it suited mostly for a gang scheduled machine. Note also that software libraries are written for several systems in Table 2. 1 to provide a standardized messaging interface, such as MPI [31] and NX [32], to the user. While it enhances program portability, this approach sacrifices efficiency for generality. In the Meiko CS 2 for example, a 10 Theta degradation is observed when an mpsc library is layered on top of its primitive mechanisms. For high performance computing, the compiler needs to ....

Paul Pierce, "The NX/2 operating system", in Proceedings of the 3rd Conference on Hypercube concurrent Computers and Applications, 1988. pp. 384--390.

....is desirable. This new standard will accomplish for parallel applications what high level languages permitted for sequential application software application programs running on different parallel machines without any changes. Numerous attempts have been made to propose a standard. Intel s NX 2 [Pierce88], PICL [Geist90] Express [Parasoft92] p4 [Butler92] and PVM [Dongarra94b] are only a few examples. Common drawbacks of these programming systems are: Designed with regard to the hardware of a particular computer manufacturer. Designed as a research project, not suitable for commercial ....

P. Pierce. "The NX/2 Operating System." Proceedings of the Third Conference on Hypercube Concurrent Computers and Applications, pages 384390. ACM Press, 1988.

....techniques. In principle we consider shared memory (SM) as well as distributed memory (DM) types of (vector) machines. More specifically, we initially concentrate on modelling parallel machines at the parallel programming interface level (i.e. DM virtual machines such as Intel iPSC alikes [14], CSTools [11] Argonne GMD macros [7] and SM virtual machines like VM EPEX [4] The Force [8] Also with respect to the range of parallel applications we hold no restrictions a priori. However, the emphasis of the actual algorithms we will consider will be on largescale numerical applications. ....

P. Pierce, "The NX/2 operating system," in Proc. 3rd Hypercube Conference, 1988.

....can be either portable to different computing platforms or tied to a particular system. PDC message passing tools written by using standard communication interfaces are usually portable but cannot fully utilize the benefits of the underlying communication network. Such tools as CMMD [48] or NX 2 [49] are specially designed to support message passing for particular systems (e.g. CMMD for CM 5 and NX 2 for Intel parallel computers) Since these tools use proprietary communication hardware and software, they perform better than the general purpose PDC message passing tools in the particular ....

P. Pierce, "The NX/2 Operating System," Proc. of the Third Conference on Hypercube Concurrent Computers and Applications, pp. 384--390, 1988.

.... A more promising starting point are the excellent synthesis of more classical models in [Ban91a] and [An91] Communication Platforms Most of the available massively parallel computers are distributed memory machine which support message passing features for communication purposes such as NX [Pie88] on the Paragon and CMMD [CMM92] on the TMC CM 5. Parallel to the commercial platforms, research centers developed portable message passing libraries such as PVM [PVM93] p4 [But92] ISIS [Bi93] or more recently the Message Passing Interface Standard MPI [MPI93] For an overview of all these ....

P. Pierce. "The NX/2 operating system". Proceedings of the Third Conference on Hypercube Concurrent Computers and Applications, ACM Press, 1988, pp. 384-390.

....that serves as an interface between the user and the computing nodes. Only the host has secondary storage, hence the computing nodes access the file system through the host. The host runs the UNIX System V operating system, and the computing node runs a proprietary operating system called the NX 2 [26] which supports, among other things, file access to the host and message passing between the nodes. The hypercube program usually consists of a host program running on the host and a node program running on every computing node. In most cases, the host program downloads the data to the node and ....

P. Pierce, "The NX/2 Operating System," in Proc. 3rd Conf. on Hypercube Concurrent Computers and Applications, Pasadena, CA, pp. 384-390, Jan. 1988.

No context found.

P. Pierce, "The NX/2 Operating System", Proceedings of the Third Conference on Hypercube Concurrent Computers and Applications, 1988, pp. 384-390.

No context found.

P. Pierce, "The NX/2 operating system," in Proceedings of 3rd Conference on Hypercube Concurrent Computers and Applications, pp. 384--390, Jan. 1988.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC