Resiliency of exascale systems has quickly become an important concern for the scientific community. Despite its importance, still much remains to be determined regarding how faults disseminate or at what rate do they impact HPC applications. The understanding of where and how fast faults propagate could lead to more efficient implementation of application-driven error detection and recovery. In this work, we propose a fault propagation framework to analyze how faults propagate in MPI applications and to understand their vulnerability to faults. We employ a combination of compiler-level code transformation and instrumentation, along with a runtime checker. Using the information provided by our framework, we employ machine learning technique to derive application fault propagation models that can be used to estimate the number of corrupted memory locations at runtime.
Aggressive CMOS technology scaling trends exacerbate the aging-related degradation of propagation delay and energy efficiency in nanoscale designs. Recently, Power-gating has been utilized as an effective low-power design technique which has also been shown to alleviate some aging impacts. However, the use of MOSFETs to realize power-gated designs will also encounter aging-induced degradations in the sleep transistors themselves which necessitates the exploration of design strategies to utilize power-gating effectively to mitigate aging. In particular, Bias Temperature Instability (BTI) which occurs during activation of power-gated voltage islands is investigated with respect to the placement of the sleep transistor in the header or footer as well as the impact of ungated input transitions on interfacial trapping. Results indicate the effectiveness of power-gating on NBTI/PBTI phenomena and propose a preferred sleep transistor configuration for maximizing higher recovery.
The strict power efficiency constraints required to achieve exascale systems will dramatically increase the number of detected and undetected transient errors in future high performance computing (HPC) systems. Among the various factors that effect system resiliency, the impact of compiler optimizations on the vulnerability of scientific applications executed on HPC systems has not been widely explored. In this work, we analyze whether and how most common compiler optimizations impact the vulnerability of several mission-critical applications, what are the trade-offs between performance and vulnerability and the causal relations between compiler optimization and application vulnerability. We show that highly-optimized code is generally more vulnerable than unoptimized code. We also show that, while increasing optimization level can drastically improve application performance as expected. However, certain cases of optimization may provide only marginal benefits, but considerably increase application vulnerability.
In this work, Field-Programmable Gate Array (FPGA) reconfigurability is exploited to realize autonomous fault recovery in mission-critical applications at runtime. The proposed Netlist-Driven Evolutionary Refurbishment technique utilizes design-time information from the circuit netlist to constrain the search space of the algorithm by up to 98.1 percent in terms of the chromosome length representing reconfigurable logic elements. This facilitates refurbishment of relatively large-sized FPGA circuits as compared to previous works. Hence, the scalability issue associated with Evolvable Hardware-Based refurbishment is addressed and improved. Experiments are conducted with multiple circuits from the MCNC benchmark suite to validate the approach and assess its benefits and limitations. Successful refurbishment of the apex4 circuit having a total of 1,252 LUTs with 10 percent spares is achieved in as few as 633 generations on average when subjected to simulated randomly injected single stuck-at faults. Moreover, the use of design-time information about the circuit undergoing refurbishment is validated as means to increase the tractability of dynamic evolvable hardware techniques.
1 January 2011 in International Journal of Reconfigurable Computing
As reconfigurable devices' capacities and the complexity of applications that use them increase, the need forself-relianceof deployed systems becomes increasingly prominent. Organic computing paradigms have been proposed for fault-tolerant systems because they promote behaviors that allow complex digital systems to adapt and survive in demanding environments. In this paper, we develop asustainable modular adaptive redundancy technique (SMART)composed of a two-layered organic system. The hardware layer is implemented on a XilinxVirtex-4Field Programmable Gate Array (FPGA) to provide self-repair using a novel approach calledreconfigurable adaptive redundancy system (RARS). The software layer supervises the organic activities on the FPGA and extends the self-healing capabilities through application-independent, intrinsic, and evolutionary repair techniques that leverage the benefits of dynamic partial reconfiguration (PR). SMART was evaluated using a Sobel edge-detection application and was shown to tolerate stressful sequences of injected transient and permanent faults while reducing dynamic power consumption by 30% compared to conventionaltriple modular redundancy (TMR)techniques, with nominal impact on the fault-tolerance capabilities. Moreover, PR is employed to keep the system on line while under repair and also to reduce repair time. Experiments have shown a 27.48% decrease in repair time when PR is employed compared to the full bitstream configuration case.