Adaptive Resource-Centric Computing (ARCC) enables a simultaneous mix of high-throughput parallel, real-time, and interactive applications through automatic discovery the correct resource assignments necessary to achieve application requirements. This approach, embodied in Tessellation manycore operating system, distributes resources QoS domains called cells. separates global decisions about allocation cells from application-specific scheduling within We examine implementation ARCC OS,...

We investigate proactive dynamic load balancing on multicore systems, in which threads are continually migrated to reduce the impact of processor/thread mismatches enhance flexibility SPMD-style programming model, and enable SPMD applications run efficiently multiprogrammed environments. present Juggle, a practical decentralized, user-space implementation balancer that emphasizes portability usability. Juggle shows performance improvements up 80% over static for UPC, OpenMP, pthreads...

Enabling message communication among concurrent computing threads without relying on mutual exclusion (i.e., locking) is highly desirable in real-time systems. This paper presents a refined version of the Non-Blocking Buffer (NBB), which lock-free interaction mechanism that enables efficient event-message between single producer thread and consumer thread. The NBB scheme presented here contains improvements over previous two aspects. First, application designers now have flexibility choosing...

The real-time specification for Java (RTSJ) offers the predictable memory management needed applications, while maintaining 's advantages of portability and ease use. RTSJ's scoped allows object lifetimes to be controlled in groups, rather than individually as C++. While easier individual lifetime management, adds programming complexity from strict rules governing access across scopes. Moreover, leaks can potentially create jitter reduce performance. To manage complexities memory, we...

To provide quick responses to users, Internet companies rely on online data systems able answer queries in milliseconds. These employ complementary overload management techniques ensure they a continued, acceptable service through-out traffic surges, where 'acceptable' partly means that serviced meet or track closely their response time objectives. Thus, this paper we present Bouncer, an admission control policy aimed keep admitted under near level objectives (SLOs) percentile times. It...

Abstract This paper presents a solution methodology for the problem of estimating distributions permeability and porosity in heterogeneous multiphase petroleum reservoirs by matching static dynamic data available. The includes, construction "fast surrogate" an objective function whose evaluation involves execution time-consuming mathematical model (i.e. reservoir numerical simulator) based on neural networks, DACE modeling, adaptive sampling. Using sampling, promising areas are searched...

Latency and cost of Internet-based services are driving the proliferation web-object caching. Memcached, most broadly deployed caching solution, is a key infrastructure component for many companies that offer via Web, such as Amazon, Facebook, Linked In, Twitter, Wikipedia, YouTube. Its aim to reduce service latency improve processing capability on back-end data servers by immutable closer client machines. Caching key-value pairs performed solely in memory. In this paper, we present novel...

Improvements in cloud-based speech recognition have led to an explosion voice assistants, as bespoke devices the home, cars, wearables or on smart phones. In this paper, we present UIVoice, through which enable assistants (that heavily utilize cloud) dynamically interact with mobile applications running edge. We a framework that can be used by third party developers easily create Voice User Interfaces (VUIs) top of existing applications. demonstrate feasibility our approach prototype based...

This paper discusses the benefits of exploiting 1) principle global-time-based coordination distributed computing actions (TCoDA) and 2) a high-level component-/object-based programming approach in developing real-time embedded software. The are discussed context concrete case study. A new major type multimedia processing applications, called Adaptive Distributed Sound Systems (ADSSs), is presented here to show compelling nature TCoDA exploitation. High-quality ADSSs impose stringent...

Virtualization allows simultaneous execution of multi-tenant workloads on the same platform, either a server or an embedded system. Unfortunately, it is non-trivial to attribute hardware events multiple virtual tenants, as some system's metrics relate whole system (e.g., RAPL energy counters). Virtualized environments have then rather incomplete picture how tenants use hardware, limiting their optimization capabilities. Thus, we propose XeM-Power, lightweight monitoring solution for Xen that...

Sources of multidimensional data are becoming more prevalent, partly due to the rise Internet Things (IoT), and so is need ingest analyze streams at rates higher than before. Some industrial IoT applications require ingesting millions records per second, while processing queries on recently ingested historical data. Unfortunately, existing database systems targeting exhibit low per-node ingestion performance, even if they can scale horizontally in distributed settings, large number nodes...

HPC facilities typically use batch scheduling to space-share jobs. In this paper we revisit time-sharing using a trace of over 2.4 million jobs obtained during 20 months operation modern petascale supercomputer. Our simulations show that produces skewed distributions with much larger slowdowns for shorter-running, jobs, whereas more uniform slowdowns. Consequently, applications strong scale, the turnaround time does not scale scheduling, but it time-sharing, resulting in turnarounds are...

Real-time CORBA is a middleware standard that has demonstrated successes in developing distributed, realtime, and embedded (DRE) systems. Customizing real-time for an application can considerably reduce the size of improve its performance. However, customizing error-prone task requires deep knowledge as well design. This paper presents ZEN-kit, graphical tool RTZen (an RTSJ-based implementation CORBA). customization achieved through modularizing so features may be inserted or removed based...

Multi-tenant virtualized infrastructures allow cloud providers to minimize costs through workload consolidation. One of the largest is power consumption, which challenging understand in heterogeneous environments. We propose a modeling methodology that tackles this complexity using divide-and-conquer approach. Our results outperform previous research work, achieving relative error 2% on average and under 4% almost all cases. Models are portable across similar architectures, enabling...

Demands have been growing in safety-critical application fields for producing networked real-time embedded computing (NREC) systems together with acceptable assurances of tight service time bounds (STBs). Here a can be defined as the amount that NREC system could take accepting request, executing an appropriate method, and returning valid result. Enabling systematic composition large-scale STB certifications has recognized highly desirable goal by research community many years. An appealing...

A hybrid approach to deriving tight execution-time bounds of program segments was proposed very recently. This symbiotically combines analytical and measurement-based methods find a execution- time bound falling between the maximum measured execution an analytically derived loose bound. It also enables estimation probability not being exceeded at run time. paper provides refined description presents techniques for measuring times acyclic-path (APSs), which are possible sequences instructions...

This article consists of a collection slides from the author's conference presentation on Tessellation operating system.