Predictive Resource Management for Next-Generation High-Performance Computing Heterogeneous Platforms

Abstract

High-Performance Computing (HPC) is rapidly moving towards the adoption of nodes characterized by an heterogeneous set of processing resources. This has already shown benefits in terms of both performance and energy efficiency. On the other side, heterogeneous systems are challenging from the application development and the resource management perspective. In this work, we discuss some outcomes of the MANGO project, showing the results of the execution of real applications on a emulated deeply heterogeneous systems for HPC. Moreover, we assessed the achievements of a proposed resource allocation policy, aiming at identifying a priori the best resource allocation options for a starting application.

TEXTAROSSA: Towards EXtreme scale Technologies and Accelerators for euROhpc hw/Sw Supercomputing Applications for exascale

“…Energy efficiency can then be guaranteed only if, on the SW side, we can rely on a suitable (hierarchical) resource management framework. Although the state-of-theart already includes some solutions, recent projects, like MANGO [28] and RECIPE [29], show that optimized solutions need to take into account the platform-specific characteristics and control knobs to profile the applications at design-time and monitor them at runtime [30], [31], enabling more accurate resource mappings [32], [33]. Furthermore, an integration of the resource manager with the programming model allows dynamically tuning the numerical accuracy (precision) of the tasks, with respect to the actual application requirements and power/energy constraints.…”

Timing Predictability in High-Performance Computing With Probabilistic Real-Time

“…The former's goal is to characterize the applications accurately and, in particular, their timing profiles. On the other hand, the latter needs to develop the knowledge of the target platform at run-time -and this is crucial for HPC platforms that are too complex to be analyzed and characterized at design-time -in order to tune the resource management policies [45]. In this regard, it is worth noting that the number of papers dealing with response-time driven resource management is only 15% of the total amount reported in the survey by Singh et al [62].…”

The RECIPE approach to challenges in deeply heterogeneous high performance systems

“…cloud-based solutions [2], potentially open to a large audience of HPC users. These emerging scenarios require that new HPC platforms are able to support multiple applications running concurrently, possibly with conflicting Quality of Service requirements [3][4][5][6][7]. In addition, at the technology level, RECIPE addresses deep heterogeneity, based on dedicated accelerators like GPUs and FPGAs, as an enabling factor for improved energy efficiency, building on the results collected from previous research projects [8,9].…”