The effectiveness of Recurrent Neural Networks (RNNs) for tasks such as Automatic Speech Recognition has fostered interest in RNN inference acceleration. Due to the recurrent nature and data dependencies of RNN computations, prior work has designed customized architectures specifically tailored to the computation pattern of RNN, getting high computation efficiency for certain chosen model sizes. However, given that the dimensionality of RNNs varies a lot for different tasks, it is crucial to generalize this efficiency to diverse configurations. In this work, we identify adaptiveness as a key feature that is missing from today’s RNN accelerators. In particular, we first show the problem of low resource utilization and low adaptiveness for the state-of-the-art RNN implementations on GPU, FPGA, and ASIC architectures. To solve these issues, we propose an intelligent tiled-based dispatching mechanism for increasing the adaptiveness of RNN computation, in order to efficiently handle the data dependencies. To do so, we propose Sharp as a hardware accelerator, which pipelines RNN computation using an effective scheduling scheme to hide most of the dependent serialization. Furthermore, Sharp employs dynamic reconfigurable architecture to adapt to the model’s characteristics. Sharp achieves 2×, 2.8×, and 82× speedups on average, considering different RNN models and resource budgets, compared to the state-of-the-art ASIC, FPGA, and GPU implementations, respectively. Furthermore, we provide significant energy reduction with respect to the previous solutions, due to the low power dissipation of Sharp (321 GFLOPS/Watt).
Continuous Vision (CV) systems are essential for emerging applications like Autonomous Driving (AD) and Augmented/Virtual Reality (AR/VR). A standard CV System-on-a-Chip (SoC) pipeline includes a frontend for image capture and a backend for executing vision algorithms. The frontend typically captures successive similar images with gradual positional and orientational variations. As a result, many regions between consecutive frames yield nearly identical results when processed in the backend. Despite this, current systems process every image region at the camera’s sampling rate, overlooking the fact that the actual rate of change in these regions could be significantly lower. In this work, we introduce δ LTA (δont’t Look Twice, it’s Alright), a novel frontend that decouples camera frame sampling from backend processing by extending the camera with the ability to discard redundant image regions before they enter subsequent CV pipeline stages. δ LTA informs the backend about the image regions that have notably changed, allowing it to focus solely on processing these distinctive areas and reusing previous results to approximate the outcome for similar ones. As a result, the backend processes each image region using different processing rates based on its temporal variation. δ LTA features a new Image Signal Processing (ISP) design providing similarity filtering functionality, seamlessly integrated with other ISP stages to incur zero-latency overhead in the worst-case scenario. It also offers an interface for frontend-backend collaboration to fine-tune similarity filtering based on the application requirements. To illustrate the benefits of this novel approach, we apply it to a state-of-the-art CV localization application, typically employed in AD and AR/VR. We show that δ LTA removes a significant fraction of unneeded frontend and backend memory accesses and redundant backend computations, which reduces the application latency by 15.22% and its energy consumption by 17%.
GPU architectures have become popular for executing general-purpose programs which rely on having a large number of threads that run concurrently to hide the latency among dependent instructions. This approach has an important cost/overhead in terms of low data locality due to the increased pressure on the memory hierarchy of the many threads being run concurrently and the extra cost of storing and managing the on-chip state of those many threads. This paper presents SOCGPU (Simple Out-of-order Core for GPU), a simple out-of-order execution mechanism that does not require register renaming nor scoreboards. It uses a small Instruction Buffer and a tiny Dependence matrix to keep track of dependencies among instructions and avoid data hazards. Evaluations for an Nvidia Tesla V100-like GPU show that SOCGPU provides a speed-up of up to 2.3 in some machine learning programs and 1.38 on average for a variety of benchmarks, while it reduces energy consumption by 6.5%, with only 2.4% area overhead.