University of Thessaly (UTH)
System software techniques to enhance reliability of modern platforms
Abstract
dc:descriptionChip manufacturers introduce redundancy at various levels of CPU design to guaranteecorrect operation even for worst-case combinations of non-idealities in process variation andsystem operating conditions. This redundancy is implemented partly in the form of wide voltagemargins.This PhD dissertation is based on the concept that these conservative design runes aremostly excessive, as they account for execution scenarios that rarely appear during the lifetime of the systems. If the faults are ignored the system will result to application crashes or even system-wide failures. Our software based approach treats these faults to enable execution in such conditions.The approach is based on the concept that many applications domains it is not the exactoutput that matters but a rough estimation of the output. Therefore, we propose a programming model in which the developer can define which parts of the application are more significant than others. The programming model extends a widely used parallel programming model, called OpenMP. The developer provides information about the significance of computations and the programming model exposes a parameter, called ratio, which can control the extend of quality degradation and energy efficiency.The idea of significance-aware computing is ported into two different computing paradigms,the fault tolerant one and the approximate. In the case of fault tolerant computing we implement a significance-centric programming model and runtime support which sets the supply voltage in a multicore CPU to sub-nominal values to reduce the energy footprint and provide mechanisms to control output quality. The developers specify the significance of application tasks respecting their contribution to the output quality and provide check and repair functions for handling faults. On a multicore system we evaluate our approach using an energy model which quantifies the energy reduction. When executing the least significant tasks unreliably, our approach leads to 20% CPU energy reduction with respect to a reliable execution and has minimal quality degradation.In the case of approximate computing, we implement a similar programming model thatpromotes the combination of the significance and ratio features. The approach using analytical models of the energy consumption of the application can iii efficiently decide the degree of approximation to meet certain user defined energy requirements.There are many application domains which require their computations to be performedwithout any errors. Our study on the x86-64 Haswell and Skylake multicore microarchitecturesreveals wide voltage which can be removed without observing any errors. These margins canreach up to 22% and 13% of the nominal supply voltage for the Skylake and Haswellarchitectures respectively. The margins vary across different microarchitectures, different chipparts of the same microarchitecture, and across different workloads.We introduce a model which can be used dynamically to adjust the supply voltage of modernmulticore x86-64 CPUs to just above the minimum required for safe operation. We identify a set of performance metrics – directly measurable via performance monitoring counters – with high predictive value for the minimum tolerable supply voltage (Vmin ). We use benchmarks that vary in terms of application domain, resource utilization and pressure, and software/hardware interaction characteristics to train a Vmin prediction model. Finally, at execution time those metrics are monitored and serve as input to the model, in order to predict and apply the appropriate Vmin for the workload. Compared to the conventional approaches, our methodology achieves up to 42% energy savings for the Skylake family and 34% for the Haswell family for complex, real-world applications.Last but not least, during the course of the thesis we implemented the infrastructure toobserve accurately the application resiliency of faults as well as to identify voltage andfrequency margins of modern processors. GemFI is a fault injection tool based on the cycleaccurate full system simulator Gem5. GemFI provides fault injection methods and is easilyextensible to support future fault models. It also supports multiple processor models and ISAsand allows fault injection in both functional and cycle-accurate simulations. GemFI offers fast-forwarding of simulation campaigns via checkpointing. Moreover, it facilitates the parallelexecution of campaign experiments on a network of workstations. XM2 enables the evaluationof software on systems operating outside their nominal margins. It supports both bare-metaland OS-controlled execution using an API to control the fault injection procedure and providesautomatic management of experimental campaigns.
Degree
thesis:*- Grantor dc:publisher
- University of Thessaly (UTH)
- Year dc:date
- 2018
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Parasyris, Konstantinos
- Παρασύρης, Κωνσταντίνος
Subjects
dc:subject × 14- Εισαγωγή σφαλμάτων
- Ενεργειακή αποδοτικότητα
- Περιθώρια τάσης
- Περιθώρια συχνότητας
- Προγραμματιστικά μοντέλα
- Fault injection
- Energy efficiency
- Voltage margins
- Frequency margins
- Programming models
- Φυσικές Επιστήμες
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Natural Sciences
- Computer and Information Sciences
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/44414