Downloads provided by UsageCounts
Este trabajo propone una extensión integral de ISA para mejorar la confiabilidad de la GPU a los efectos transitorios. Se proponen, implementan y combinan tres instrucciones adicionales con la duplicación de rutas de datos basada en software. Los códigos de programa modificados se comparan con las técnicas de tolerancia a fallos basadas en software de última generación en términos de tiempo de ejecución. El área del circuito se evalúa con respecto a la arquitectura original de la GPU y se realiza una campaña de inyección de fallos para evaluar la fiabilidad. Los resultados muestran que esta extensión integral de ISA mejora el rendimiento y las capacidades de detección de fallas de los enfoques basados en software a costos insignificantes en términos de área de circuito. Este trabajo puede ayudar a los ingenieros a diseñar arquitecturas de GPU más eficientes y resistentes.
Ce travail propose une extension ISA complète pour améliorer la fiabilité du GPU aux effets transitoires. Trois instructions supplémentaires sont proposées, mises en œuvre et combinées à une duplication logicielle du chemin de données. Les codes de programme modifiés sont comparés aux techniques logicielles de pointe de tolérance aux pannes en termes de temps d'exécution. La zone du circuit est évaluée par rapport à l'architecture GPU d'origine, et une campagne d'injection de défauts est effectuée pour évaluer la fiabilité. Les résultats montrent que cette extension ISA complète améliore les performances et les capacités de détection des pannes des approches logicielles à des coûts négligeables en termes de surface de circuit. Ce travail peut aider les ingénieurs à concevoir des architectures de GPU plus efficaces et résilientes.
This work proposes a comprehensive ISA extension to improve GPU reliability to transient effects. Three additional instructions are proposed, implemented, and combined with software-based datapath duplication. Modified program codes are compared to state-of-the-art software-based fault tolerance techniques in terms of execution time. The circuit area is evaluated against the original GPU architecture, and a fault injection campaign is performed to assess reliability. Results show that this comprehensive ISA extension improves performance and fault detection capabilities of software-based approaches at negligible costs in terms of circuit area. This work can help engineers in designing more efficient and resilient GPU architectures.
يقترح هذا العمل تمديدًا شاملاً لمعايير التدقيق الدولية لتحسين موثوقية وحدة معالجة الرسومات للتأثيرات العابرة. يتم اقتراح ثلاث تعليمات إضافية وتنفيذها ودمجها مع ازدواجية مسار البيانات القائم على البرامج. تتم مقارنة رموز البرنامج المعدلة بأحدث تقنيات التسامح مع الأخطاء القائمة على البرامج من حيث وقت التنفيذ. يتم تقييم منطقة الدائرة مقابل بنية وحدة معالجة الرسومات الأصلية، ويتم إجراء حملة حقن الأعطال لتقييم الموثوقية. تُظهر النتائج أن هذا التمديد الشامل لمعايير المحاسبة الدولية يحسن الأداء وقدرات اكتشاف الأعطال للنهج القائمة على البرامج بتكاليف لا تذكر من حيث منطقة الدائرة. يمكن أن يساعد هذا العمل المهندسين في تصميم هياكل وحدات معالجة الرسومات أكثر كفاءة ومرونة.
software-based hardening, Parallel computing, Instruction set, Datapath, GPUs, Computer Networks and Communications, Fault Tolerance, Quantum mechanics, Reliability engineering, Engineering, Fault injection, FOS: Electrical engineering, electronic engineering, information engineering, Parallel Computing and Performance Optimization, Register file, GPU Computing, Computer architecture, Electrical and Electronic Engineering, Embedded system, Extension (predicate logic), Performance Optimization, Physics, In-field hardening, Fault tolerance, Reliability Evaluation, Power (physics), Computer science, Programming language, Operating system, Fault Tolerance in Electronic Systems, Reliability (semiconductor), Hardware and Architecture, Physical Sciences, Computer Science, Transient Faults, Distributed Fault Tolerance and Consistency in Systems, Software
software-based hardening, Parallel computing, Instruction set, Datapath, GPUs, Computer Networks and Communications, Fault Tolerance, Quantum mechanics, Reliability engineering, Engineering, Fault injection, FOS: Electrical engineering, electronic engineering, information engineering, Parallel Computing and Performance Optimization, Register file, GPU Computing, Computer architecture, Electrical and Electronic Engineering, Embedded system, Extension (predicate logic), Performance Optimization, Physics, In-field hardening, Fault tolerance, Reliability Evaluation, Power (physics), Computer science, Programming language, Operating system, Fault Tolerance in Electronic Systems, Reliability (semiconductor), Hardware and Architecture, Physical Sciences, Computer Science, Transient Faults, Distributed Fault Tolerance and Consistency in Systems, Software
| selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | 10 | |
| popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network. | Top 10% | |
| influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | Top 10% | |
| impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network. | Top 10% |
| views | 5 | |
| downloads | 10 |

Views provided by UsageCounts
Downloads provided by UsageCounts