arXiv subjects

Behzad Salami

Publications and source records attributed to Behzad Salami.

23 records · Page 2Linked to original sources

arXiv2019

On the Resilience of Deep Learning for Reduced-voltage FPGAs

Deep Neural Networks (DNNs) are inherently computation-intensive and also power-hungry. Hardware accelerators such as Field Programmable Gate Arrays (FPGAs) are a promising solution that can satisfy these requirements for both embedded and High-Performance Computing (HPC) systems. In FPGAs, as well as CPUs and GPUs, aggressive voltage scaling below the nominal level is an effective technique for power dissipation minimization. Unfortunately, bit-flip faults start to appear as the voltage is scaled down closer to the transistor threshold due to timing issues, thus creating a resilience issue. This paper experimentally evaluates the resilience of the training phase of DNNs in the presence of voltage underscaling related faults of FPGAs, especially in on-chip memories. Toward this goal, we have experimentally evaluated the resilience of LeNet-5 and also a specially designed network for CIFAR-10 dataset with different activation functions of Rectified Linear Unit (Relu) and Hyperbolic Tangent (Tanh). We have found that modern FPGAs are robust enough in extremely low-voltage levels and that low-voltage related faults can be automatically masked within the training iterations, so there is no need for costly software- or hardware-oriented fault mitigation techniques like ECC. Approximately 10% more training iterations are needed to fill the gap in the accuracy. This observation is the result of the relatively low rate of undervolting faults, i.e., <0.1\%, measured on real FPGA fabrics. We have also increased the fault rate significantly for the LeNet-5 network by randomly generated fault injection campaigns and observed that the training accuracy starts to degrade. When the fault rate increases, the network with Tanh activation function outperforms the one with Relu in terms of accuracy, e.g., when the fault rate is 30% the accuracy difference is 4.92%.

cs.LG↗

arXiv2019

A Novel FPGA-Based High Throughput Accelerator For Binary Search Trees

This paper presents a deeply pipelined and massively parallel Binary Search Tree (BST) accelerator for Field Programmable Gate Arrays (FPGAs). Our design relies on the extremely parallel on-chip memory, or Block RAMs (BRAMs) architecture of FPGAs. To achieve significant throughput for the search operation on BST, we present several novel mechanisms including tree duplication as well as horizontal, duplicated, and hybrid (horizontal-vertical) tree partitioning. Also, we present efficient techniques to decrease the stalling rates that can occur during the parallel tree search. By combining these techniques and implementations on Xilinx Virtex-7 VC709 platform, we achieve up to 8X throughput improvement gain in comparison to the baseline implementation, i.e., a fully-pipelined FPGA-based accelerator.

cs.DC↗

arXiv2019

Hardware Versus Software Fault Injection of Modern Undervolted SRAMs

To improve power efficiency, researchers are experimenting with dynamically adjusting the supply voltage of systems below the nominal operating points. However, production systems are typically not allowed to function on voltage settings that is below the reliable limit. Consequently, existing software fault tolerance studies are based on fault models, which inject faults on random fault locations using fault injection techniques. In this work we study whether random fault injection is accurate to simulate the behavior of undervolted SRAMs. Our study extends the Gem5 simulator to support fault injection on the caches of the simulated system. The fault injection framework uses fault maps, which describe the faulty bits of SRAMs, as inputs. To compare random fault injection and hardware guided fault injection, we use two types of fault maps. The first type of maps are created through undervolting real SRAMs and observing the location of the erroneous bits, whereas the second type of maps are created by corrupting random bits of the SRAMs. During our study we corrupt the L1-Dcache of the simulated system and we monitor the behavior of the two types of fault maps on the resiliency of six benchmarks. The difference among the resiliency of a benchmark when tested with the different fault maps can be up to 24%.

cs.PF↗

arXiv2019

Evaluating Built-in ECC of FPGA on-chip Memories for the Mitigation of Undervolting Faults

Voltage underscaling below the nominal level is an effective solution for improving energy efficiency in digital circuits, e.g., Field Programmable Gate Arrays (FPGAs). However, further undervolting below a safe voltage level and without accompanying frequency scaling leads to timing related faults, potentially undermining the energy savings. Through experimental voltage underscaling studies on commercial FPGAs, we observed that the rate of these faults exponentially increases for on-chip memories, or Block RAMs (BRAMs). To mitigate these faults, we evaluated the efficiency of the built-in Error-Correction Code (ECC) and observed that more than 90% of the faults are correctable and further 7% are detectable (but not correctable). This efficiency is the result of the single-bit type of these faults, which are then effectively covered by the Single-Error Correction and Double-Error Detection (SECDED) design of the built-in ECC. Finally, motivated by the above experimental observations, we evaluated an FPGA-based Neural Network (NN) accelerator under low-voltage operations, while built-in ECC is leveraged to mitigate undervolting faults and thus, prevent NN significant accuracy loss. In consequence, we achieve 40% of the BRAM power saving through undervolting below the minimum safe voltage level, with a negligible NN accuracy loss, thanks to the substantial fault coverage by the built-in ECC.

cs.AR↗

arXiv2018

On the Resilience of RTL NN Accelerators: Fault Characterization and Mitigation

Machine Learning (ML) is making a strong resurgence in tune with the massive generation of unstructured data which in turn requires massive computational resources. Due to the inherently compute- and power-intensive structure of Neural Networks (NNs), hardware accelerators emerge as a promising solution. However, with technology node scaling below 10nm, hardware accelerators become more susceptible to faults, which in turn can impact the NN accuracy. In this paper, we study the resilience aspects of Register-Transfer Level (RTL) model of NN accelerators, in particular, fault characterization and mitigation. By following a High-Level Synthesis (HLS) approach, first, we characterize the vulnerability of various components of RTL NN. We observed that the severity of faults depends on both i) application-level specifications, i.e., NN data (inputs, weights, or intermediate), NN layers, and NN activation functions, and ii) architectural-level specifications, i.e., data representation model and the parallelism degree of the underlying accelerator. Second, motivated by characterization results, we present a low-overhead fault mitigation technique that can efficiently correct bit flips, by 47.3% better than state-of-the-art methods.

cs.LG↗