Carlo Cavazzoni

Unveiling Eurora — Thermal and power characterization of the most energy-efficient supercomputer in the world

Eurora (EURopean many integrated cORe Architecture) is today the most energy efficient supercomputer in the world. Ranked 1st in the Green500 in July 2013, is a prototype built from Eurotech and Cineca toward next-generation Tier-0 systems in the PRACE 2IP EU project. Euroras outstanding energy-efficiency is achieved by adopting a direct liquid cooling solution and a heterogeneous architecture with best-in-class general purpose HW components (Intel Xeon E5, Intel Xeon Phi and NVIDIA Kepler K20). In this paper we present a novel, low-overhead monitoring infrastructure capable to track in detail and in real-time the thermal and power characteristics of Euroras components with fine-grained resolution. Our experiments give insights on Euroras thermal/power trade-offs and highlight opportunities for run-time power/thermal management and optimization.

LiGen: a high performance workflow for chemistry driven de novo design.

Tools for molecular de novo design are actively sought incorporating sets of chemical rules for fast and efficient identification of structurally new chemotypes endowed with a desired set of biological properties. In this paper, we present LiGen, a suite of programs which can be used sequentially or as stand-alone tools for specific purposes. In its standard application, LiGen modules are used to define input constraints, either structure-based, through active site identification, or ligand-based, through pharmacophore definition, to docking and to de novo generation. Alternatively, individual modules can be combined in a user-defined manner to generate project-centric workflows. Specific features of LiGen are the use of a pharmacophore-based docking procedure which allows flexible docking without conformer enumeration and accurate and flexible reactant mapping coupled with reactant tagging through substructure searching. The full description of LiGen functionalities is presented.

The ANTAREX approach to autotuning and adaptivity for energy efficient HPC systems

The ANTAREX project aims at expressing the application self-adaptivity through a Domain Specific Language (DSL) and to runtime manage and autotune applications for green and heterogeneous High Performance Computing (HPC) systems up to Exascale. The DSL approach allows the definition of energy-efficiency, performance, and adaptivity strategies as well as their enforcement at runtime through application autotuning and resource and power management. We show through a mini-app extracted from one of the project application use cases some initial exploration of application precision tuning by means enabled by the DSL.

Use of experimental design to optimize docking performance: the case of LiGenDock, the docking module of LiGen, a new de novo design program.

On route toward a novel de novo design program, called LiGen, we developed a docking program, LiGenDock, based on pharmacophore models of binding sites, including a non-enumerative docking algorithm. In this paper, we present the functionalities of LiGenDock and its accompanying module LiGenPocket, aimed at the binding site analysis and structure-based pharmacophore definition. We also report the optimization procedure we have carried out to improve the cognate docking and virtual screening performance of LiGenDock. In particular, we applied the design of experiments (DoE) methodology to screen the set of user-adjustable parameters to identify those having the largest influence on the accuracy of the results (which ensure the best performance in pose prediction and in virtual screening approaches) and then to choose their optimal values. The results are also compared with those obtained by two popular docking programs, namely, Glide and AutoDock for pose prediction, and Glide and DOCK6 for Virtual Screening.

Continuous learning of HPC infrastructure models using big data analytics and in-memory processing tools

Exascale computing represents the next leap in the HPC race. Reaching this level of performance is subject to several engineering challenges such as energy consumption, equipment-cooling, reliability and massive parallelism. Model-based optimization is an essential tool in the design process and control of energy efficient, reliable and thermally constrained systems. However, in the Exascale domain, model learning techniques tailored to the specific supercomputer require real measurements and must therefore handle and analyze a massive amount of data coming from the HPC monitoring infrastructure. This becomes rapidly a “big data” scale problem. The common approach where measurements are first stored in large databases and then processed is no more affordable due to the increasingly storage costs and lack of real-time support. Nowadays instead, cloud-based machine learning techniques aim to build on-line models using real-time approaches such as “stream processing” and “in-memory” computing, that avoid storage costs and enable fastdata processing. Moreover, the fast delivery and adaptation of the models to the quick data variations, make the decision stage of the optimization loop more effective and reliable. In this paper we leverage scalable, lightweight and flexible IoT technologies, such as the MQTT protocol, to build a highly scalable HPC monitoring infrastructure able to handle the massive sensor data produced by next-gen HPC components. We then show how state-of-the art tools for big data computing and analysis, such as Apache Spark, can be used to manage the huge amount of data delivered by the monitoring layer and to build adaptive models in real-time using on-line machine learning techniques.

Autotuning and adaptivity approach for energy efficient Exascale HPC systems: The ANTAREX approach

The main goal of the ANTAREX 1 project is to express by a Domain Specific Language (DSL) the application self-adaptivity and to runtime manage and autotune applications for green and heterogeneous High Performance Computing (HPC) systems up to the Exascale level. Key innovations of the project include the introduction of a separation of concerns between self-adaptivity strategies and application functionalities. The DSL approach will allow the definition of energy-efficiency, performance, and adaptivity strategies as well as their enforcement at runtime through application autotuning and resource and power management.

ANTAREX -- AutoTuning and Adaptivity appRoach for Energy Efficient eXascale HPC Systems

The main goal of the ANTAREX project is to express by a Domain Specific Language (DSL) the application self-adaptivity and to runtime manage and autotune applications for green and heterogeneous High Performance Computing (HPC) systems up to the Exascale level. Key innovations of the project include the introduction of a separation of concerns between self-adaptivity strategies and application functionalities. The DSL approach will allow the definition of energy-efficiency, performance, and adaptivity strategies as well as their enforcement at runtime through application autotuning and resource and power management.

Autotuning and adaptivity in energy efficient HPC systems: the ANTAREX toolbox

Designing and optimizing applications for energy-efficient High Performance Computing systems up to the Exascale era is an extremely challenging problem. This paper presents the toolbox developed in the ANTAREX European project for autotuning and adaptivity in energy efficient HPC systems. In particular, the modules of the ANTAREX toolbox are described as well as some preliminary results of the application to two target use cases. 1

Quantifying the Impact of Variability and Heterogeneity on the Energy Efficiency for a Next-Generation Ultra-Green Supercomputer

Supercomputers, nowadays, aggregate a large number of nodes featuring the same nominal HW components (e.g., processors and GPGPUS). In real-life machines, the chips populating each node are subject to a wide range of variability sources, related to performance and temperature operating points (i.e., ACPI p-states) as well as process variations and die binning. Eurora is a fully operational supercomputer prototype that topped July 2013 Green500 and it represents a unique ’living lab’ for next-generation ultra-green supercomputers. In this paper we evaluate and quantify the impact of variability on Euroras energy-performance tradeoffs under a wide range of workloads intensity. Our experiments demonstrate that variability comes from hardware component mismatches as well as from the interplay between run-time energy management and workload variations. Thus, variability has a significant impact on energy efficiency even at the moderate scale of the Eurora machine, thereby substantiating the critical importance of variability management in future green supercomputers.

Cooling-aware node-level task allocation for next-generation green HPC systems

Energy-efficiency is of primary interest in future HPC systems as their computational growth is limited by the supercomputer peak power consumption. A significant part of the power consumed by a supercomputer machine is caused by the cooling infrastructure. Todays thermal design is based on coarse grain models which consider the silicon die of the processing elements as an isothermal surface. Similarly feedback control loops uses the same assumption to modulate the cooling effort with the goal of reducing cooling cost and maintaining the silicon temperature in a safe working range. Recent processors development has brought into the market CPUs that integrate a large number of complex cores. Differently from massively parallel CPUs for which the area and power consumption of each core is very limited, the cores of these processors can consume tens of watts and thus, under heterogeneous workloads, creating significant thermal gradients. In this paper we first characterize the power and thermal characteristics of new server-class Intel Xeon computing node based on Haswell v3 architecture considering both the computational and the cooling components. We show that these systems are characterized by significant on-die thermal gradients and that the current O.S. task allocation strategy is not capable of taking advantage of that, leading to max CPU temperature and extra cooling activity. To solve this issue we propose a novel task allocation strategy that reduces the cooling power while matching the HPC performance requirements.