Tight temporal bounds for dataflow applications mapped onto shared resources

Author(s):  
Hadi Alizadeh Ara ◽  
Marc Geilen ◽  
Twan Basten ◽  
Amir Behrouzian ◽  
Martijn Hendriks ◽  
...  
Keyword(s):  
2012 ◽  
Author(s):  
Desiree Fullemann ◽  
Rebecca Brauchli ◽  
Gregor Jenny ◽  
Georg Bauer
Keyword(s):  

2020 ◽  
Vol 638 ◽  
pp. 149-164
Author(s):  
GM Svendsen ◽  
M Ocampo Reinaldo ◽  
MA Romero ◽  
G Williams ◽  
A Magurran ◽  
...  

With the unprecedented rate of biodiversity change in the world today, understanding how diversity gradients are maintained at mesoscales is a key challenge. Drawing on information provided by 3 comprehensive fishery surveys (conducted in different years but in the same season and with the same sampling design), we used boosted regression tree (BRT) models in order to relate spatial patterns of α-diversity in a demersal fish assemblage to environmental variables in the San Matias Gulf (Patagonia, Argentina). We found that, over a 4 yr period, persistent diversity gradients of species richness and probability of an interspecific encounter (PIE) were shaped by 3 main environmental gradients: bottom depth, connectivity with the open ocean, and proximity to a thermal front. The 2 main patterns we observed were: a monotonic increase in PIE with proximity to fronts, which had a stronger effect at greater depths; and an increase in PIE when closer to the open ocean (a ‘bay effect’ pattern). The originality of this work resides on the identification of high-resolution gradients in local, demersal assemblages driven by static and dynamic environmental gradients in a mesoscale seascape. The maintenance of environmental gradients, specifically those associated with shared resources and connectivity with an open system, may be key to understanding community stability.


2020 ◽  
Author(s):  
Nicholas Harp ◽  
Michael D. Dodd ◽  
Maital Neta

Cognitive resources are needed for successful executive functioning; when resources are limited due to competing demands, task performance is impaired. Although some tasks are accomplished with relatively few resources (e.g., judging trustworthiness and emotion in others), others are more complex. Specifically, in the face of emotional ambiguity (i.e., stimuli that do not convey a clear positive or negative meaning, such as a surprised facial expression), our decisions to approach or avoid appear to rely on the availability of top-down regulatory resources to overcome an initial negativity bias. Cognition-emotion interaction theories (e.g., dual competition) posit that emotion and executive processing rely on shared resources, suggesting that competing demands would hamper these regulatory responses towards emotional ambiguity. Here, we employed a 2x2 design to investigate the effects of load (low versus high) and domain (non-emotional vs. emotional) on evaluations of surprised faces. As predicted, there were domain-specific effects, such that categorizations of surprise were more negative for emotional than non-emotional loads. Consistent with prior work, low load (regardless of domain; i.e., domain-general) was associated with greater response competition on trials resulting in a positive categorization, showing that positive categorizations are characterized by an initial negativity. This effect was diminished under high load. These results lend insight into the resources supporting a positive valence bias by demonstrating that emotion-specific regulatory resources are important for overriding the initial negativity in response to emotional ambiguity. However, both domain-general and domain-specific loads impact the underlying processes.


2020 ◽  
pp. 239965442096524
Author(s):  
Mariska JM Bottema ◽  
Simon R Bush ◽  
Peter Oosterveer

The Thai aquaculture sector faces a range of production, market and financial risks that extend beyond the private space of farms to include public spaces and shared resources. The Thai state has attempted to manage these shared risks through its Plang Yai (or ‘Big Area’) agricultural extension program. Using the lens of territorialization, this paper investigates how, through the Plang Yai program, risk management is institutionalized through spatially explicit forms of collaboration amongst farmers and between farmers and (non-)state actors. We focus on how four key policy instruments brought together under Plang Yai delimited multiple territories of risk management over shrimp and tilapia production in Chantaburi and Chonburi provinces. Our findings demonstrate how these policy instruments address risks through dissimilar but overlapping territories that are selectively biased toward facilitating the individual management of production risks, whilst enabling both the individual and collective management of market and financial risks. This raises questions about the suitability of addressing aquaculture risks by controlling farmer behavior through state-led designation of singular, spatially explicit areas. The findings also indicate the multiple roles of the state in territorializing risk management, providing a high degree of flexibility, which is especially valuable in landscapes shared by many users, connected to (global) value chains and facing diverse risks. In doing so we demonstrate that understanding the territorialization of production landscapes in a globalizing world requires a dynamic approach recognizing the multiplicity of territories that emerge in risk management processes.


Author(s):  
Umar Ibrahim Minhas ◽  
Roger Woods ◽  
Georgios Karakonstantis

AbstractWhilst FPGAs have been used in cloud ecosystems, it is still extremely challenging to achieve high compute density when mapping heterogeneous multi-tasks on shared resources at runtime. This work addresses this by treating the FPGA resource as a service and employing multi-task processing at the high level, design space exploration and static off-line partitioning in order to allow more efficient mapping of heterogeneous tasks onto the FPGA. In addition, a new, comprehensive runtime functional simulator is used to evaluate the effect of various spatial and temporal constraints on both the existing and new approaches when varying system design parameters. A comprehensive suite of real high performance computing tasks was implemented on a Nallatech 385 FPGA card and show that our approach can provide on average 2.9 × and 2.3 × higher system throughput for compute and mixed intensity tasks, while 0.2 × lower for memory intensive tasks due to external memory access latency and bandwidth limitations. The work has been extended by introducing a novel scheduling scheme to enhance temporal utilization of resources when using the proposed approach. Additional results for large queues of mixed intensity tasks (compute and memory) show that the proposed partitioning and scheduling approach can provide higher than 3 × system speedup over previous schemes.


2020 ◽  
Vol 13 (1) ◽  
Author(s):  
Hendrik De Smet

AbstractOne view of language change sees changes as originating in erroneous deviations from the linguistic norm and diffusing through social transmission. An alternative is to see changes as originating in speakers’ problem-solving activities and spreading in response to system pressures that reflect speakers’ recurrent communicative needs and shared resources. It is argued here that the latter may well be the dominant scenario.


2021 ◽  
Vol 18 (3) ◽  
pp. 1-25
Author(s):  
Weijia Song ◽  
Christina Delimitrou ◽  
Zhiming Shen ◽  
Robbert Van Renesse ◽  
Hakim Weatherspoon ◽  
...  

Infrastructure-as-a-Service cloud providers sell virtual machines that are only specified in terms of number of CPU cores, amount of memory, and I/O throughput. Performance-critical aspects such as cache sizes and memory latency are missing or reported in ways that make them hard to compare across cloud providers. It is difficult for users to adapt their application’s behavior to the available resources. In this work, we aim to increase the visibility that cloud users have into shared resources on public clouds. Specifically, we present CacheInspector , a lightweight runtime that determines the performance and allocated capacity of shared caches on multi-tenant public clouds. We validate CacheInspector ’s accuracy in a controlled environment, and use it to study the characteristics and variability of cache resources in the cloud, across time, instances, availability regions, and cloud providers. We show that CacheInspector ’s output allows cloud users to tailor their application’s behavior, including their output quality, to avoid suboptimal performance when resources are scarce.


Sign in / Sign up

Export Citation Format

Share Document