Difference between time shared and space shared algorithms

Question

I just came across this term "time shared and space shared algorithms". I couldn't get much information about it in the web. Can anyone shed some light on these terms, specially on space sharing?

Answer 1

In my understanding, the terms time-sharing and space-sharing refer to the way the resources of a machine are shared among computational tasks. For instance, several processes typically take turns in accessing the CPU. This is time-sharing, because over a period of time, we observe more than one process utilizing the resource. On the contrary, when we assign cores of a CPU to individual processes, we observe more than one process only by looking at more than one core, thus the term space-sharing. A definition from [1]:

Space-sharing: The machine may be partitioned into sets of processors (clusters). Each cluster is allocated to a single job that is allowed to run to completion (RTC).

Time-sharing: More than one job may be allocated to a cluster, in which case each job runs for some quantum of time before being preempted to allow other jobs to run.

A more recent, but less broad definition of the terms can be found in the context of a Cloud environment [2]:

assign specific CPU cores to specific VMs (a space-shared policy), dynamically distribute the capacity of a core among VMs (time-shared policy)

Both definitions refer to exclusive access (until a logical task is done) to as space-sharing, while interwoven access to a resource (during a logical task) is referred to as time-sharing.

[1] A. B. Downey, “Predicting Queue Times on Space-Sharing Parallel Computers.,” IPPS, pp. 209–218, 1997.

[2] R. N. Calheiros, R. Ranjan, A. Beloglazov, C. A. F. De Rose, and R. Buyya, “CloudSim: a toolkit for modeling and simulation of cloud computing environments and evaluation of resource provisioning algorithms,” SPE, vol. 41, no. 1, pp. 23–50, Aug. 2010.