In order to provide shared memory in large-scale multiprocessors, techniques to hide the latency of shared memory accesses must be developed. In this paper, we describe the latency hiding mechanisms employed by the Ga...
In order to provide shared memory in large-scale multiprocessors, techniques to hide the latency of shared memory accesses must be developed. In this paper, we describe the latency hiding mechanisms employed by the Galactica Net scalable distributed shared memory architecture being developed at the center for highperformancecomputing. We introduce our novel technique for maintaining the coherence of shared data caches, based on a flexible hardware-supported but software-controlled mechanism supporting both update and invalidate based protocols. We also consider the use of alternative memory consistency models, and find that the use of weaker consistency models is an effective way to hide memory reference latency in the Galactica Net architecture. Preliminary performance evaluations indicate that together these mechanisms are able to hide a significant amount of the memory reference latency, thus increasing the scalability of the architecture.
The class of NUMA (nongniform Memory Access time) shared memory architectures is becoming increasingly important with the desire for larger scale multiprocessors. Insuchmachines, the placement and movement of code and...
详细信息
This paper looks at the problem of ensuring performance of real-time applications hosted on Galactica Net∗, a mesh-based distributed cache coherent shared memory multiprocessing system. A method for determining strict...
详细信息
Mixing of particles by chaotic flow fields was simulated on the Connection Machine. We assigned each cell to the processor and kept the coordinates of particles residing on the cell in the local memory of the processo...
详细信息
Steady-state and time-dependent two-dimensional thermal convection in a Boussinesq, infinite-Prandtl-number fluid with stress-free boundaries has been investigated. Two independent numerical methods have been employed...
Steady-state and time-dependent two-dimensional thermal convection in a Boussinesq, infinite-Prandtl-number fluid with stress-free boundaries has been investigated. Two independent numerical methods have been employed to calculate the evolution of convective flows in a rectangular box with aspect ratio λ=1.8 in a Rayleigh-number (Ra) range of 106109. With increasing Ra, greater than 107, the flow reveals the presence of disconnected thermals, rather than connected plumes, driven by a persistent large-scale circulation. Such features have also been reported from laboratory convection experiments in the regime of hard turbulence. Extensive calculations were performed (up to 140 overturns) in order to reach the statistically stationary regime for strongly chaotic flows. A Gaussian distribution with a mean value Nut was derived from the time history of the Nusselt (Nu) numbers. The value of Nut can be directly obtained by solving the steady-state equations via an iteration procedure. Thus the stationary flow obtained from the steady-state method resembles the turbulent flow in a statistical sense. Since the iteration procedure is about 104 times faster than calculating the full time-dependent evolution, it allows for the systematic investigation of the heat-transfer Nu-Ra relationship and other types of scaling laws. The steady-state and time-dependent experiments indicate that a power-law exponent of β=0.315 holds for the Nu-Ra relation for stress-free boundaries in the entire range of Ra. No indication of a jump in the exponent was found in the transition to hard turbulence.
This paper looks at the problem of ensuring performance of real-time applications hosted on Galactica Net, a mesh-based distributed cache coherent shared memory multiprocessing system. A method for determining strict ...
详细信息
This paper looks at the problem of ensuring performance of real-time applications hosted on Galactica Net, a mesh-based distributed cache coherent shared memory multiprocessing system. A method for determining strict upper bounds on worst case latencies in wormhole routed meshes is presented, and it is shown that the update latency of Galactica Net is deterministic. A tool for determining strict upper bounds for shared memory update latencies has been developed so that differing real-time process placements in a Galactica Net system may be compared to minimize update latency bounds.< >
The use of characteristics-based methods for the advection-dominated regimes in thermal convection is investigated. An operator-splitting method applied to the advection-diffusion equation for the very large Peclet (P...
The use of characteristics-based methods for the advection-dominated regimes in thermal convection is investigated. An operator-splitting method applied to the advection-diffusion equation for the very large Peclet (Pe) number regime is presented. In this approach two partial differential equations representing both the purely hyperbolic and the parabolic components must be solved simultaneously. This method has been compared with (1) the Galerkin approximation, (2) the streamwise upwinding Petrov-Galerkin method, and (3) the characteristics-based method using the Lagrangian formulation for the time-derivative operator of the advection-diffusion equation. Solution accuracy of the operator-splitting method improves with larger Pe, while the accuracy of other methods deteriorates with Pe. For the nonlinear problem of two-dimensional thermal convection the Lagrangian method is found to be most computationally efficient. With this Lagrangian method, time-dependent, thermal convection solutions of extremely high Rayleigh number (Ra), up to 3 X 10(9), for infinite Prandtl number are obtained. For an aspect ratio of 1.8 the exponent in the scaling of the Nusselt number (Nu) with Ra in time-dependent convection is determined to be 0.301 in the hard turbulent regime and is smaller than in the soft turbulent regime. The behavior of this exponent as a consequence of the transition to hard turbulence agrees with experimental findings. Horizontal Fourier spectra of the thermal fields outside the boundary layers reveal a transition in the high wave-number domain from 1/k to 1/k2 in the transition from soft to hard turbulent regimes. Analysis of the kinetic energy spectra E(k) shows an asymptotic decay of E(k) close to K-2, for large k, spanning over two decades in wave number for strongly time-dependent convection.
This book constitutes the refereed proceedings of the 10th International Workshop on Ophthalmic Medical Image Analysis, OMIA 2023, held in conjunction with the 26th International Conference on Medical Imaging and Comp...
详细信息
ISBN:
(数字)9783031440137
ISBN:
(纸本)9783031440120
This book constitutes the refereed proceedings of the 10th International Workshop on Ophthalmic Medical Image Analysis, OMIA 2023, held in conjunction with the 26th International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2023, in Vancouver, Canada, in October 2023.
ThisvolumeispublishedastheproceedingsoftheRussian-GermanAdvanced Research workshop on Computational Science and highperformance C- puting in Novosibirsk Academgorodok in September 2003. The contributions of these pro...
详细信息
ISBN:
(数字)9783540323761
ISBN:
(纸本)9783540241201;9783642438103
ThisvolumeispublishedastheproceedingsoftheRussian-GermanAdvanced Research workshop on Computational Science and highperformance C- puting in Novosibirsk Academgorodok in September 2003. The contributions of these proceedings were provided and edited by the authors, chosen after a careful selection and reviewing. The workshop was organized by the institute of Computational Techno- gies SB RAS (Novosibirsk, Russia) and the highperformancecomputingcenter Stuttgart (Stuttgart, Germany). The objective was the discussion of the latest results in computational science and to develop a close coope- tion between Russian and German specialists in the above-mentioned ?eld. The main directions of the workshop are associated with the problems of computational hydrodynamics, application of mathematical methods to the development of new generation of materials, environment protection pr- lems, development of algorithms, software and hardware support for hi- performance computation, and designing modern facilities for visualization of computational modelling results. The importance of the workshop topics was con?rmed by the partici- tion of representatives of major research organizations engaged in the so- tion of the most complex problems of mathematical modelling, development of new algorithms, programs and key elements of new information techno- gies. Among the Russian participants were researchers of the institutes of the Siberian Branch of the Russian Academy of Sciences: institute of Com- tational Technologies, institute of Computational Mathematics and Mat- matical Geophysics, institute of Computational Modelling, Russian Federal Nuclear center, All-Russian Research institute of Experimental Physics, - merovo State University.
暂无评论