Wyniki wyszukiwania - Biblioteka Nauki

Nowa wersja platformy, zawierająca wyłącznie zasoby pełnotekstowe, jest już dostępna.
Przejdź na https://bibliotekanauki.pl

Ograniczanie wyników

Znaleziono wyników: 3

Liczba wyników na stronie

Wyniki wyszukiwania

Sortuj według:

Ogranicz wyniki do:

Improving energy efficiency of supercomputer systems through software-aided liquid cooling management

100%

Januszewski R. , Różycki R. , Waligóra G.

tom Vol. 43, No. 2

89--103

Many fields of modern science rely more and more on the immense computing power of supercomputers. Modern, multi-thousand node systems can consume megawatts of electrical energy in highly uneven manner, challenging the data center infrastructure, both power and cooling coils. The traditional way of managing the infrastructure makes each subsystem of a data center (e.g. cooling) independent from all other in the way it relies only on local sensors to manage the infrastructure. The erratic nature of computing in a large data center makes this approach suboptimal. In the paper we show that by challenging the traditional split between the infrastructure and the computing equipment, one can gain significant boost in energy efficiency of the entire ecosystem. A solution that predicts cooling power demand basing on the information from a supercomputer resource manager, and then sets up the parameters of the cooling loop, is presented along with potential benefits in terms of reduction of the power draw.

Discrete-continuous project scheduling with preemptable activities

100%

Różycki R. , Waligóra G. , Węglarz J.

2016

tom Vol. 64, nr 2

383--393

In this paper, discrete-continuous project scheduling problems with preemptable activities are considered. In these problems, activities of a project simultaneously require discrete and continuous resources for their execution. The activities are preemptable, and the processing rate of each activity is a continuous, increasing function of the amount of a single continuous resource allotted to the activity at a time. The problem is to find a precedence- and discrete resource-feasible schedule and, simultaneously, continuous resource allocation that would minimize the project duration. Convex and concave processing rate functions are considered separately. We show that for convex functions the problem is simple, whereas for concave functions a special methodology has to be developed. We discuss the methodology for three cases of the problem: no discrete resource constraints, one discrete resource being a set of parallel, identical machines, and an arbitrary number of discrete resources. In each case we analyze separately independent and precedence-related activities. Some conclusions and directions for future research are given.

Scheduling battery charging jobs with linearly decreasing power demands to minimize the total time

100%

Różycki R. , Waligóra G. , Węglarz J.

Bulletin of the Polish Academy of Sciences. Technical Sciences

2020

tom Vol. 68, nr 2

299--306

In this work we consider a problem from the field of power- and energy-aware scheduling, in which a set of batteries have to be charged in a minimum time. The formulated problem is to schedule independent and nonpreemptable jobs to minimize the schedule length, where each job requires some amount of power and consumes a certain amount of energy during its processing. We assume that the power demand of each job linearly decreases with time, as it is the case when Li-ion batteries are being charged. For the assumed job model we prove that each next job should be started as soon as the required amount of power is available. Basing on the proven theorem we formulate a procedure generating a minimum-length schedule for an assumed order of jobs. We also analyze the case of identical jobs, and show some interesting properties of this case.