Section: New Results


Simgrid is a toolkit providing core functionalities for the simulation of distributed applications in heterogeneous distributed environments. Although it was initially designed to study large distributed computing environments such as grids, we have recently applied it to performance prediction of HPC configurations.

  • Indeed, multi-core architectures comprising several GPUs have become mainstream but obtaining the maximum performance of such heterogeneous machines is challenging as it requires to carefully offload computations and manage data movements between the different processing units. The most promising and successful approaches so far build on task-based runtimes that abstract the machine and rely on opportunistic scheduling algorithms. As a consequence, the problem gets shifted to choosing the task granularity, task graph structure, and optimizing the scheduling strategies. Trying different combinations of these different alternatives is also itself a challenge. Indeed, getting accurate measurements requires reserving the target system for the whole duration of experiments. Furthermore, observations are limited to the few available systems at hand and may be difficult to generalize. In [21] , we show how we crafted a coarse-grain hybrid simulation/emulation of StarPU, a dynamic runtime for hybrid architectures, over SimGrid. This approach allows to obtain performance predictions of classical dense linear algebra kernels accurate within a few percents and in a matter of seconds, which allows both runtime and application designers to quickly decide which optimization to enable or whether it is worth investing in higher-end GPUs or not. Additionally, it allows to conduct robust and extensive scheduling studies in a controlled environment whose characteristics are very close to real platforms while having reproducible behavior. In [30] , we have extended this approach to the simulation of a multithreaded multifrontal QR solver of sparse matrices: QR-MUMPS. In our approach, the target high-end machines are calibrated only once to derive sound performance models. These models can then be used at will to quickly predict and study in a reproducible way the performance of such irregular and resource-demanding applications using solely a commodity laptop. Our approach also allows to study the memory consumption along time, which is a critical factor for such applications.

  • Beside the inherent heterogeneity of distributed computing infrastructures, storage is also a essential component to cope with the tremendous increase in scientific data production and the ever-growing need for data analysis and preservation. Understanding the performance of a storage subsystem or dimensioning it properly is an important concern for which simulation can help. In [29] , we detail how we have extended SimGrid with storage simulation capacities and we list several concrete use cases of storage simulations in clusters, grids, clouds, and data centers for which the proposed extension would be beneficial.

Ψ2is a simulation software of markovian models that is able to provide a perfect sampling of the stationary distribution. In [12] , we consider open Jackson networks with losses with mixed finite and infinite queues and analyze the efficiency of sampling from their exact stationary distribution. We show that perfect sampling is possible, although the underlying Markov chain may have an infinite state space. The main idea is to use a Jackson network with infinite buffers (that has a product form stationary distribution) to bound the number of initial conditions to be considered in the coupling from the past scheme. We also provide bounds on the sampling time of this new perfect sampling algorithm for acyclic or hyper-stable networks. These bounds show that the new algorithm is considerably more efficient than existing perfect samplers even in the case where all queues are finite. We illustrate this efficiency through numerical experiments. We also extend our approach to variable service times and non-monotone networks such as queueing networks with negative customers.