Reliability Evaluation of Mixed-Precision Architectures

Santos, Fernando Fernandes dos; Lunardi, Caio; Oliveira, Daniel; Libano, F.; Rech, Paolo

doi:10.1109/hpca.2019.00041

Cited by 27 publications

(14 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Quick- For Volta GPU, we also perform the analysis for FIT rates of codes executed with different precisions: Double 64bit (D), Float 32bit (F), and Half 16bit (H). A higher precision functional unit has a higher area and a higher probability of being hit by a neutron since the circuit area is bigger [1], [32]. Thus, higher precision implies more bits to store data, which is linearly dependent on the FIT rate.…”

Section: Neutron Beam Experiments Resultsmentioning

confidence: 99%

Experimental Findings on the Sources of Detected Unrecoverable Errors in GPUs

Santos

Malde

Cazzaniga

et al. 2022

IEEE Trans. Nucl. Sci.

Self Cite

View full text Add to dashboard Cite

We investigate the sources of Detected Unrecoverable Errors (DUEs) in Graphics Processing Units (GPU) exposed to a neutron beam. Illegal memory accesses and interface errors are among the more likely sources of DUEs. ECC increases the launch failure events. Our test procedure has shown that ECC can reduce the DUEs caused by Illegal Address access up to 92% for Kepler and up to 98% for Volta. Additionally, we analyze if the compiler optimizations can impact the DUE sources distribution for the Matrix Multiplication. We found that the machine codes generated by the different optimization levels can change the DUE source by no more than 24% on average.

show abstract

Section: Neutron Beam Experiments Resultsmentioning

confidence: 99%

Experimental Findings on the Sources of Detected Unrecoverable Errors in GPUs

Santos

Malde

Cazzaniga

et al. 2022

IEEE Trans. Nucl. Sci.

Self Cite

View full text Add to dashboard Cite

show abstract

“…SDCs are particularly critical, as their silent nature makes them extremely hard to detect and they can potentially lead to unstable or unknown system states. Previous studies have already evaluated the reliability of CPUs, ARM, heterogeneous devices, FPGAs, GPUs, and Xeon Phi through radiation experiments [4], [11], fault-injection [12], [13], or both [14]- [16].…”

Section: They Can Induce the Following Effects On The Program Outputmentioning

confidence: 99%

“…Recently, the reliability of mixed-precision architectures has been evaluated through fault injection and radiation beam experiments [11]. Preliminary results show that, for GPUs, the higher the precision, the higher the expected error rate.…”

Section: B Approximate Computing and Mixed-precisionmentioning

confidence: 99%

Reduced Precision DWC: An Efficient Hardening Strategy for Mixed-Precision Architectures

Santos

Brandalero

Sullivan

et al. 2022

IEEE Trans. Comput.

Self Cite

View full text Add to dashboard Cite

Duplication with Comparison (DWC) is an effective software-level solution to improve the reliability of computing devices. However, it introduces performance and energy consumption overheads that could be unsuitable for high-performance computing or real-time safety-critical applications.In this work, we present Reduced-Precision Duplication with Comparison (RP-DWC) as a means to lower the overhead of DWC by executing the redundant copy in reduced precision. RP-DWC is particularly suitable for modern mixed-precision architectures, such as NVIDIA GPUs, that feature dedicated functional units for computing with programmable accuracy. We discuss the benefits and challenges associated with RP-DWC and show that the intrinsic difference between the mixed-precision copies allows for detecting most, but not all, errors. However, as the undetected faults are the ones that fall into the difference between precisions, they are the ones that produce a much smaller impact on the application output and, thus, might be tolerated.We investigate RP-DWC impact into fault detection, performance, and energy consumption on Volta GPUs. Through fault injection and beam experiment, using three microbenchmarks and four real applications, we show that RP-DWC achieves an excellent coverage (up to 86%) with minimal overheads (as low as 0.1% time and 24% energy consumption overhead)

show abstract

“…Since the existing fault injection tools targeting GPU programs lack regional fault analysis [13,14,16,19,45], we design and implement a regional fault injection tool for our purpose. While the general-purpose fault injectors target random fault injection points, our tool introduces faults in a specific code portion (source code lines) and examines the execution of the program after injection time.…”

Section: Regional Fault Injection Toolmentioning

confidence: 99%

“…Santos et al [16] evaluate the soft error vulnerability of mixed-precision architectures and utilize a fault injection tool implemented on cuda-gdb debugger for target GPU devices. In their other work [42], Santos et al evaluate the soft error vulnerability of the object detection frameworks by utilizing CAROL-FI and investigate Histogram of Oriented Gradients (HOG) and You Only Look Once (YOLO) benchmarks to examine kernel and layer vulnerabilities.…”

Section: Related Workmentioning

confidence: 99%

Regional soft error vulnerability and error propagation analysis for GPGPU applications

Karadaş

2021

J Supercomput

View full text Add to dashboard Cite

The wide use of GPUs for general-purpose computations as well as graphics programs makes soft errors a critical concern. Evaluating the soft error vulnerability of GPGPU programs and employing efficient fault tolerance techniques for more reliable execution become more important. Protecting only the most error-sensitive program regions maintains an acceptable reliability level by eliminating the large performance overheads due to redundant operations. Therefore, fine-grained regional soft error vulnerability analysis is crucial for the systems targeting both performance and reliability. In this work, we present a regional fault injection framework and perform a detailed error propagation analysis to evaluate the soft error vulnerability of GPGPU applications. We evaluate both intra-kernel and inter-kernel vulnerabilities for a set of programs and quantify the severity of the data corruptions by considering metrics other than SDC rates. Our experimental study demonstrates that the code regions inside GPGPU programs exhibit different characteristics in terms of soft error vulnerability and the soft errors corrupting the variables propagate into the program output in several ways. We present the potential impact of our analysis by discussing the usage scenarios after we compile our observations acquired from our empirical work.

show abstract

Reliability Evaluation of Mixed-Precision Architectures

Cited by 27 publications

References 31 publications

Experimental Findings on the Sources of Detected Unrecoverable Errors in GPUs

Experimental Findings on the Sources of Detected Unrecoverable Errors in GPUs

Reduced Precision DWC: An Efficient Hardening Strategy for Mixed-Precision Architectures

Regional soft error vulnerability and error propagation analysis for GPGPU applications

Contact Info

Product

Resources

About