extreme performance
Recently Published Documents


TOTAL DOCUMENTS

60
(FIVE YEARS 13)

H-INDEX

10
(FIVE YEARS 2)

2021 ◽  
Vol 55 (1) ◽  
pp. 88-98
Author(s):  
Mohammed Islam Naas ◽  
François Trahay ◽  
Alexis Colin ◽  
Pierre Olivier ◽  
Stéphane Rubini ◽  
...  

Tracing is a popular method for evaluating, investigating, and modeling the performance of today's storage systems. Tracing has become crucial with the increase in complexity of modern storage applications/systems, that are manipulating an ever-increasing amount of data and are subject to extreme performance requirements. There exists many tracing tools focusing either on the user-level or the kernel-level, however we observe the lack of a unified tracer targeting both levels: this prevents a comprehensive understanding of modern applications' storage performance profiles. In this paper, we present EZIOTracer, a unified I/O tracer for both (Linux) kernel and user spaces, targeting data intensive applications. EZIOTracer is composed of a userland as well as a kernel space tracer, complemented with a trace analysis framework able to merge the output of the two tracers, and in particular to relate user-level events to kernel-level ones, and vice-versa. On the kernel side, EZIOTracer relies on eBPF to offer safe, low-overhead, low memory footprint, and flexible tracing capabilities. We demonstrate using FIO benchmark the ability of EZIOTracer to track down I/O performance issues by relating events recorded at both the kernel and user levels. We show that this can be achieved with a relatively low overhead that ranges from 2% to 26% depending on the I/O intensity.


2021 ◽  
Author(s):  
Nicole D. Montijn ◽  
Lotte Gerritsen ◽  
Iris. M. Engelhard

ABSTRACTTrauma memories can appear dissociated from their original temporal context, and are often relived as they occur in the here-and-now. Potentially these temporal distortions already occur during encoding of the aversive experience as a consequence of stress. Here, 86 participants were subjected to either a stress or control induction, after which they learned the temporal structure of four virtual days. In these virtual days, time was scaled and participants could use clock cues to construe the passage of time within a day. We examined whether stress causes a shift in the learning strategy from one based on virtual time to one based on event sequence. Our results do not show a discernible impact of stress on memory for temporal context, in terms of both sequence memory and more fine-grained representations of time. The stress groups showed more extreme performance trajectories, either good or poor, across all measures. However, as time estimations were overall quite poor it is unclear to what extent this reflected a true strategy shift. Future avenues of research that can build on these findings are discussed.


2020 ◽  
Vol 11 (1) ◽  
pp. 20190116 ◽  
Author(s):  
Christos Kotsalos ◽  
Jonas Latt ◽  
Joel Beny ◽  
Bastien Chopard

We propose a highly versatile computational framework for the simulation of cellular blood flow focusing on extreme performance without compromising accuracy or complexity. The tool couples the lattice Boltzmann solver Palabos for the simulation of blood plasma, a novel finite-element method (FEM) solver for the resolution of deformable blood cells, and an immersed boundary method for the coupling of the two phases. The design of the tool supports hybrid CPU–GPU executions (fluid, fluid–solid interaction on CPUs, deformable bodies on GPUs), and is non-intrusive, as each of the three components can be replaced in a modular way. The FEM-based kernel for solid dynamics outperforms other FEM solvers and its performance is comparable to state-of-the-art mass–spring systems. We perform an exhaustive performance analysis on Piz Daint at the Swiss National Supercomputing Centre and provide case studies focused on platelet transport, implicitly validating the accuracy of our tool. The tests show that this versatile framework combines unprecedented accuracy with massive performance, rendering it suitable for upcoming exascale architectures.


2020 ◽  
Vol 13 (3) ◽  
pp. 1293-1316
Author(s):  
Andreas Hönl ◽  
Philip Meissner ◽  
Torsten Wulf

AbstractCore self-evaluation summarizes a decision maker’s self-worth. This key personality trait has been shown to lead to extreme performance consequences of either winning or losing big. We suggest that these extreme performance outcomes may partly rest in how core self-evaluation affects executive’s perception and evaluation of risk in choices under uncertainty. We conducted a choice experiment building on the original prospect theory experiments with 97 executives, in which we measured the effect of core self-evaluation on risk behavior. As a robustness test, we replicated and validated our findings with a larger sample of 111 executives. Building on the tenets of prospect theory, we show that decision makers with high levels of core self-evaluation are less loss averse. Surprisingly, this effect differs depending on whether gains or losses are highlighted in the decision. For gains, higher levels of core self-evaluation are associated with behaviors that are closer to risk neutrality. For losses, however, we find that higher levels of core self-evaluation further enhance the risk-seeking behavior of decision makers. These findings contribute towards understanding the effects of core self-evaluation in the work environment as well as in the decision process and provide an additional lens for studying how the personality of executives affects choices under uncertainty.


2020 ◽  
Vol 287 (1935) ◽  
pp. 20201578
Author(s):  
Allyn Nguyen ◽  
Jordan P. Balaban ◽  
Emanuel Azizi ◽  
Robert J. Talmadge ◽  
A. Kristopher Lappin

The southern alligator lizard ( Elgaria multicarinata ) exhibits a courtship behaviour during which the male firmly grips the female's head in his jaws for many hours at a time. This extreme behaviour counters the conventional wisdom that reptilian muscle is incapable of powering high-endurance behaviours. We conducted in situ experiments in which the jaw-adductor muscles of lizards were stimulated directly while bite force was measured simultaneously. Fatigue tests were performed by stimulating the muscles with a series of tetanic trains. Our results show that a substantial sustained force gradually develops during the fatigue test. This sustained force persists after peak tetanic forces have declined to a fraction of their initial magnitude. The observed sustained force during in situ fatigue tests is consistent with the courtship behaviour of these lizards and probably reflects physiological specialization. The results of molecular analysis reveal that the jaw muscles contain masticatory and tonic myosin fibres. We propose that the presence of tonic fibres may explain the unusual sustained force properties during mate-holding behaviour. The characterization of muscle properties that facilitate extreme performance during specialized behaviours may reveal general mechanisms of muscle function, especially when done in light of convergently evolved systems exhibiting similar performance characteristics.


2020 ◽  
Vol 117 (19) ◽  
pp. 10445-10454 ◽  
Author(s):  
Stephen M. Deban ◽  
Jeffrey A. Scales ◽  
Segall V. Bloom ◽  
Charlotte M. Easterling ◽  
Mary Kate O’Donnell ◽  
...  

The evolution of ballistic tongue projection in plethodontid salamanders—a high-performance and thermally robust musculoskeletal system—is ideal for examining how the components required for extreme performance in animal movement are assembled in evolution. Our comparative data on whole-organism performance measured across a range of temperatures and the musculoskeletal morphology of the tongue apparatus were examined in a phylogenetic framework and combined with data on muscle contractile physiology and neural control. Our analysis reveals that relatively minor evolutionary changes in morphology and neural control have transformed a muscle-powered system with modest performance and high thermal sensitivity into a spring-powered system with extreme performance and functional robustness in the face of evolutionarily conserved muscle contractile physiology. Furthermore, these changes have occurred in parallel in both major clades of this largest family of salamanders. We also find that high-performance tongue projection that exceeds available muscle power and thermal robustness of performance coevolve, both being emergent properties of the same elastic-recoil mechanism. Among the taxa examined, we find muscle-powered and fully fledged elastic systems with enormous performance differences, but no intermediate forms, suggesting that incipient elastic mechanisms do not persist in evolutionary time. A growing body of data from other elastic systems suggests that similar coevolution of traits may be found in other ectothermic animals with high performance, particularly those for which thermoregulation is challenging or ecologically costly.


2020 ◽  
Author(s):  
Arnau Folch ◽  
Josep de la Puente ◽  
Laura Sandri ◽  
Benedikt Halldórsson ◽  
Andreas Fichtner ◽  
...  

<p>The Center of Excellence for Exascale in Solid Earth (ChEESE; https://cheese-coe.eu) is promoting the use of upcoming Exascale and extreme performance computing capabilities in the area of Solid Earth by harnessing institutions in charge of operational monitoring networks, tier-0 supercomputing centers, academia, hardware developers and third parties from SMEs, Industry and public-governance. The scientific challenging ambition is to prepare 10 European open-source flagship codes to solve Exascale problems on computational seismology, magnetohydrodynamics, physical volcanology, tsunamis, and data analysis. Preparation to Exascale is considering code inter-kernel aspects of simulation workflows like data management and sharing following the FAIR principles, I/O, post-process and visualization. The project is articulated around 12 Pilot Demonstrators (PDs) in which flagship codes are used for near real-time seismic simulations and full-wave inversion, ensemble-based volcanic ash dispersal forecasts, faster than real-time tsunami simulations and physics-based hazard assessments for earthquakes, volcanoes and tsunamis. This is a first step towards enabling of operational e-services requiring of extreme HPC on urgent computing, early warning forecast of geohazards, hazard assessment and data analytics. Additionally, and in collaboration with the European Plate Observing System (EPOS), ChEESE will promote and facilitate the integration of HPC services to widen the access to codes and fostering transfer of know-how to Solid Earth user communities. In this regard, the project aims at acting as a hub to foster HPC across the Solid Earth Community and related stakeholders and to provide specialized training on services and capacity building measures.</p>


Sign in / Sign up

Export Citation Format

Share Document