Impact of Soft Errors on Large-Scale FPGA Cloud Computing

Abstract

FPGAs are being used in large numbers within cloud computing to provide high-performance, low-power alternatives to more traditional computing structures. While FPGAs provide a number of important benefits to cloud computing environments, they are susceptible to radiation-induced soft errors, which can lead to silent data corruption or system instability. Although soft errors within a single FPGA occur infrequently, soft errors in large-scale FPGAs systems can occur at a relatively high rate. This paper investigates the failure rate of several FPGA applications running within an FPGA cloud computing node by performing fault injection experiments to determine the susceptibility of these applications to soft-errors. The results from these experiments suggest that silent data corruption will occur every few hours within a 100,000 node FPGA system and that such a system can only maintain high-levels of reliability for short periods of operation. These results suggest that soft-error detection and mitigation techniques may be needed in large-scale FPGA systems.

Simulation and empirical validation of new sensitivity based reliability analysis technique for processors deployed in industrial drives

“…The most commonly occurring sensitive faults of the processor are SEU, Silent Data Corruption (SDC) and Single Event Transient (SET). The SDC is not detectable through the application, randomly occurs through the design and passes incorrect data (Keller and Wirthlin, 2019). SEU corrupts memory values related to devising configuration and active design state in general.…”

MRNG: Accessing Cosmic Radiation as an Entropy Source for a Non-Deterministic Random Number Generator

“…The effect on SEUs within large-scale Field Programmable Gate Array (FPGAs) systems was discussed by [ 55 ]; they found that an SEU can be expected every 3.75 h within a 100,000 FPGAs system. Such errors most likely caused the acceleration problem within Toyota vehicles [ 56 , 57 ].…”

Evaluation of Xilinx Deep Learning Processing Unit under Neutron Irradiation

“…As a result, datacenter market leaders increasingly integrate FPGA System-on-Chip (SoC) in their infrastructure to target complex AI applications. Companies like Microsoft, Amazon Web Services, and Baidu scale up AI and high-performance applications on hundreds of thousands of Intel and Xilinx FPGA devices [1].…”