اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدEDEN: Enabling Energy-Efficient, High-Performance Deep Neural Network Inference Using Approximate DRAM
460
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
The effectiveness of deep neural networks (DNN) in vision, speech, and language processing has prompted a tremendous demand for energy-efficient high-performance DNN inference systems. Due to the increasing memory intensity of most DNN workloads, main memory can dominate the systems energy consumption and stall time. One effective way to reduce the energy consumption and increase the performance of DNN inference systems is by using approximate memory, which operates with reduced supply voltage and reduced access latency parameters that violate standard specifications. Using approximate memory reduces reliability, leading to higher bit error rates. Fortunately, neural networks have an intrinsic capacity to tolerate increased bit errors. This can enable energy-efficient and high-performance neural network inference using approximate DRAM devices. Based on this observation, we propose EDEN, a general framework that reduces DNN energy consumption and DNN evaluation latency by using approximate DRAM devices, while strictly meeting a user-specified target DNN accuracy. EDEN relies on two key ideas: 1) retraining the DNN for a target approximate DRAM device to increase the DNNs error tolerance, and 2) efficient mapping of the error tolerance of each individual DNN data type to a corresponding approximate DRAM partition in a way that meets the user-specified DNN accuracy requirements. We evaluate EDEN on multi-core CPUs, GPUs, and DNN accelerators with error models obtained from real approximate DRAM devices. For a target accuracy within 1% of the original DNN, our results show that EDEN enables 1) an average DRAM energy reduction of 21%, 37%, 31%, and 32% in CPU, GPU, and two DNN accelerator architectures, respectively, across a variety of DNNs, and 2) an average (maximum) speedup of 8% (17%) and 2.7% (5.5%) in CPU and GPU architectures, respectively, when evaluating latency-bound DNNs.
قيم البحث
اقرأ أيضاً
This paper presents GPU performance optimization and scaling results for inference models of the Sparse Deep Neural Network Challenge 2020. Demands for network quality have increased rapidly, pushing the size and thus the memory requirements of many
neural networks beyond the capacity of available accelerators. Sparse deep neural networks (SpDNN) have shown promise for reining in the memory footprint of large neural networks. However, there is room for improvement in implementing SpDNN operations on GPUs. This work presents optimized sparse matrix multiplication kernels fused with the ReLU function. The optimized kernels reuse input feature maps from the shared memory and sparse weights from registers. For multi-GPU parallelism, our SpDNN implementation duplicates weights and statically partition the feature maps across GPUs. Results for the challenge benchmarks show that the proposed kernel design and multi-GPU parallelization achieve up to 180 tera-edges per second inference throughput. These results are up to 4.3x faster for a single GPU and an order of magnitude faster at full scale than those of the champion of the 2019 Sparse Deep Neural Network Graph Challenge for the same generation of NVIDIA V100 GPUs. Using the same implementation, we also show single-GPU throughput on NVIDIA A100 is 2.37$times$ faster than V100.
Deep learning models typically use single-precision (FP32) floating point data types for representing activations and weights, but a slew of recent research work has shown that computations with reduced-precision data types (FP16, 16-bit integers, 8-
bit integers or even 4- or 2-bit integers) are enough to achieve same accuracy as FP32 and are much more efficient. Therefore, we designed fbgemm, a high-performance kernel library, from ground up to perform high-performance quantized inference on current generation CPUs. fbgemm achieves efficiency by fusing common quantization operations with a high-performance gemm implementation and by shape- and size-specific kernel code generation at runtime. The library has been deployed at Facebook, where it delivers greater than 2x performance gains with respect to our current production baseline.
Recently published methods enable training of bitwise neural networks which allow reduced representation of down to a single bit per weight. We present a method that exploits ensemble decisions based on multiple stochastically sampled network models
to increase performance figures of bitwise neural networks in terms of classification accuracy at inference. Our experiments with the CIFAR-10 and GTSRB datasets show that the performance of such network ensembles surpasses the performance of the high-precision base model. With this technique we achieve 5.81% best classification error on CIFAR-10 test set using bitwise networks. Concerning inference on embedded systems we evaluate these bitwise networks using a hardware efficient stochastic rounding procedure. Our work contributes to efficient embedded bitwise neural networks.
We propose all-optical neural networks characterized by very high energy efficiency and performance density of inference. We argue that the use of microcavity exciton-polaritons allows to take advantage of the properties of both photons and electrons
in a seamless manner. This results in strong optical nonlinearity without the use of optoelectronic conversion. We propose a design of a realistic neural network and estimate energy cost to be at the level of attojoules per bit, also when including the optoelectronic conversion at the input and output of the network, several orders of magnitude below state-of-the-art hardware implementations. We propose two kinds of nonlinear binarized nodes based either on optical phase shifts and interferometry or on polariton spin rotations.
State-of-the-art convolutional neural networks (CNNs) yield record-breaking predictive performance, yet at the cost of high-energy-consumption inference, that prohibits their widely deployments in resource-constrained Internet of Things (IoT) applica
tions. We propose a dual dynamic inference (DDI) framework that highlights the following aspects: 1) we integrate both input-dependent and resource-dependent dynamic inference mechanisms under a unified framework in order to fit the varying IoT resource requirements in practice. DDI is able to both constantly suppress unnecessary costs for easy samples, and to halt inference for all samples to meet hard resource constraints enforced; 2) we propose a flexible multi-grained learning to skip (MGL2S) approach for input-dependent inference which allows simultaneous layer-wise and channel-wise skipping; 3) we extend DDI to complex CNN backbones such as DenseNet and show that DDI can be applied towards optimizing any specific resource goals including inference latency or energy cost. Extensive experiments demonstrate the superior inference accuracy-resource trade-off achieved by DDI, as well as the flexibility to control such trade-offs compared to existing peer methods. Specifically, DDI can achieve up to 4 times computational savings with the same or even higher accuracy as compared to existing competitive baselines.
الأسئلة المقترحة
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
