MétaCan
Menu
Back to cohort

Improving time predictability of shared hardware resources in real-time multicore systems : emphasis on the space domain

2016· dissertation· en· W2561968680 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

fundA Canadian funder is recorded on the work.
no affNo Canadian affiliation: this work is invisible to an affiliation-only frame.
No Canadian affiliation. An affiliation-only frame, the usual design, would never have seen this work. It is one of the works that make the case for inverting the frame.

Bibliographic record

Venuenot available
Typedissertation
Languageen
FieldComputer Science
TopicReal-Time Systems Scheduling
Canadian institutionsnot available
FundersMinisterio de Ciencia e InnovaciónUniversità degli Studi di PadovaEuropean Space AgencyBarcelona Supercomputing CenterNunavut General Monitoring Plan
KeywordsComposabilityComputer sciencePredictabilityMulti-core processorStatic timing analysisEmbedded systemComponent (thermodynamics)CorrectnessDistributed computingWorst-case execution timeProcess (computing)Computer architectureExecution timeParallel computingOperating system

Abstract

fetched live from OpenAlex

Critical Real-Time Embedded Systems (CRTES) follow a verification and validation process on the timing and functional correctness. This process includes the timing analysis that provides Worst-Case Execution Time (WCET) estimates to provide evidence that the execution time of the system, or parts of it, remain within the deadlines. A key design principle for CRTES is the incremental qualification, whereby each software component can be subject to verification and validation independently of any other component, with obvious benefits for cost. At timing level, this requires time composability, such that the timing behavior of a function is not affected by other functions. CRTES are experiencing an unprecedented growth with rising performance demands that have motivated the use of multicore architectures. Multicores can provide the performance required and bring the potential of integrating several software functions onto the same hardware. However, multicore contention in the access to shared hardware resources creates a dependence of the execution time of a task with the rest of the tasks running simultaneously. This dependence threatens time predictability and jeopardizes time composability. In this thesis we analyze and propose hardware solutions to be applied on current multicore designs for CRTES to improve time predictability and time composability, focusing on the on-chip bus and the memory controller. At hardware level, we propose new bus and memory controller designs that control and mitigate contention between different cores and allow to have time composability by design, also in the context of mixed-criticality systems. At analysis level, we propose contention prediction models that factor the impact of contenders and don¿t need modifications to the hardware. We also propose a set of Performance Monitoring Counters (PMC) that provide evidence about the contention. We give an special emphasis on the Space domain focusing on the Cobham Gaisler NGMP multicore processor, which is currently assessed by the European Space Agency for its future missions. Los Sistemas Críticos Empotrados de Tiempo Real (CRTES) siguen un proceso de verificación y validación para su correctitud funcional y temporal. Este proceso incluye el análisis temporal que proporciona estimaciones de el peor caso del tiempo de ejecución (WCET) para dar evidencia de que el tiempo de ejecución del sistema, o partes de él, permanecen dentro de los límites temporales. Un principio de diseño clave para los CRTES es la cualificación incremental, por la que cada componente de software puede ser verificado y validado independientemente del resto de componentes, con beneficios obvios para el coste. A nivel temporal, esto requiere composabilidad temporal, por la que el comportamiento temporal de una función no se ve afectado por otras funciones. CRTES están experimentando un crecimiento sin precedentes con crecientes demandas de rendimiento que han motivado el uso the arquitecturas multi-núcleo (multicore). Los procesadores multi-núcleo pueden proporcionar el rendimiento requerido y tienen el potencial de integrar varias funcionalidades software en el mismo hardware. A pesar de ello, la interferencia entre los diferentes núcleos que aparece en los recursos compartidos de os procesadores multi núcleo crea una dependencia del tiempo de ejecución de una tarea con el resto de tareas ejecutándose simultáneamente en el procesador. Esta dependencia amenaza la predictabilidad temporal y compromete la composabilidad temporal. En esta tésis analizamos y proponemos soluciones hardware para ser aplicadas en los diseños multi núcleo actuales para CRTES que mejoran la predictabilidad y composabilidad temporal, centrándose en el bus y el controlador de memoria internos al chip. A nivel de hardware, proponemos nuevos diseños de buses y controladores de memoria que controlan y mitigan la interferencia entre los diferentes núcleos y permiten tener composabilidad temporal por diseño, también en el contexto de sistemas de criticalidad mixta. A nivel de análisis, proponemos modelos de predicción de la interferencia que factorizan el impacto de los núcleos y no necesitan modificaciones hardware. También proponemos un conjunto de Contadores de Control del Rendimiento (PMC) que proporcionoan evidencia de la interferencia. En esta tésis, damós especial importancia al dominio espacial, centrándonos en el procesador mutli núcleo Cobham Gaisler NGMP, que está siendo actualmente evaluado por la Agencia Espacial Europea para sus futuras misiones.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.003
metaresearch head score (Gemma)0.001
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMeta-epidemiology (narrow)
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Bench or experimental · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.738
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0030.001
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0010.000
Bibliometrics0.0000.001
Science and technology studies0.0000.000
Scholarly communication0.0000.001
Open science0.0020.000
Research integrity0.0010.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.011
GPT teacher head0.238
Teacher spread0.226 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Quick stats

Citations0
Published2016
Admission routes1
Has abstractyes

Explore more

Same topicReal-Time Systems SchedulingFrench-language works237,207