An official website of the United States government
In recent years, machine learning research has largely shifted focus from the cloud to the edge. While the resulting algorithm- and hardware-level optimizations have enabled local execution for the majority of deep neural networks (DNNs) on edge devices, the sheer magnitude of DNNs associated with real-time video detection workloads has forced them to remain relegated to remote execution in the cloud. This problematic when combined with the strict latency requirements that are coupled with these workloads, and imposes a unique set of challenges not directly addressed in prior works. In this work, we design MobiEye, a cloud-based video detection system optimized for deployment in real-time mobile applications. MobiEye is able to achieve up to a 32% reduction in latency when compared to a conventional implementation of video detection system with only a marginal reduction in accuracy.
more »
« less
EdgePC: Efficient Deep Learning Analytics for Point Clouds on Edge Devices
Recently, point cloud (PC) has gained popularity in modeling various 3D objects (including both synthetic and real-life) and has been extensively utilized in a wide range of applications such as AR/VR, 3D reconstruction, and autonomous driving. For such applications, it is critical to analyze/understand the surrounding scenes properly. To achieve this, deep learning based methods (e.g., convolutional neural networks (CNNs)) have been widely employed for higher accuracy. Unlike the deep learning on conventional 2D images/videos, where the feature computation (matrix multiplication) is the major bottleneck, in point cloud-based CNNs, the sample and neighbor search stages are the primary bottlenecks, and collectively contribute to 54% (up to 80%) of the overall execution latency on a typical edge device. While prior efforts have attempted to solve this issue by designing custom ASICs or pipelining the neighbor search with other stages, to our knowledge, none of them has tried to “structurize” the unstructured PC data for improving computational efficiency. In this paper, we first explore the opportunities of structurizing PC data using Morton code (which is originally designed to map data from a high dimensional space to one dimension, while preserving spatial locality) and observe that there is a huge scope to “skip” the sample and neighbor search computation by operating on the “structurized” PC data. Based on this, we propose two approximation techniques for the sampling and neighbor search stages. We implemented our proposals on an NVIDIA Jetson AGX Xavier edge GPU board. The evaluation results collected on six different workloads show that our design can accelerate the sample and neighbor search stages by 3.68× (up to 5.21×) with minimal impact on inference accuracy. This acceleration in turn results in 1.55× speedup in the end-to-end execution latency and saves 33% of energy expenditure.
more »
« less
- Award ID(s):
- 1763681
- PAR ID:
- 10440142
- Date Published:
- Journal Name:
- International Symposium on Computer Architecture 2023
- Page Range / eLocation ID:
- 1 to 14
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Point cloud computation has become an increasingly more important workload thanks to its applications in autonomous driving. Unlike dense 2D computation, point cloud convolution has sparse and irregular computation patterns and thus requires dedicated inference system support with specialized high-performance kernels. While existing point cloud deep learning libraries have developed different dataflows for convolution on point clouds, they assume a single dataflow throughout the execution of the entire model. In this work, we systematically analyze and improve existing dataflows. Our resulting system, TorchSparse++, achieves 2.9x, 3.3x, 2.2x and 1.8x measured end-to-end speedup on an NVIDIA A100 GPU over the state-of-the-art MinkowskiEngine, SpConv 1.2, TorchSparse and SpConv v2 in inference respectively. Furthermore, TorchSparse++ is the only system to date that supports all necessary primitives for 3D segmentation, detection, and reconstruction workloads in autonomous driving. Code is publicly released at https://github.com/mit-han-lab/torchsparse.more » « less
-
Point cloud is an important type of geometric data structure for many embedded applications such as autonomous driving and augmented reality. Current Point Cloud Networks (PCNs) have proven to achieve great success in using inference to perform point cloud analysis, including object part segmentation, shape classification, and so on. However, point cloud applications on the computing edge require more than just the inference step. They require an end-to-end (E2E) processing of the point cloud workloads: pre-processing of raw data, input preparation, and inference to perform point cloud analysis. Current PCN approaches to support end-to-end processing of point cloud workload cannot meet the real-time latency requirement on the edge, i.e., the ability of the AI service to keep up with the speed of raw data generation by 3D sensors. Latency for end-to-end processing of the point cloud workloads stems from two reasons: memory-intensive down-sampling in the pre-processing phase and the data structuring step for input preparation in the inference phase. In this paper, we present HgPCN, an end-to-end heterogeneous architecture for real-time embedded point cloud applications. In HgPCN, we introduce two novel methodologies based on spatial indexing to address the two identified bottlenecks. In the Pre-processing Engine of HgPCN, an Octree-Indexed-Sampling method is used to optimize the memory-intensive down-sampling bottleneck of the pre-processing phase. In the Inference Engine, HgPCN extends a commercial DLA with a customized Data Structuring Unit which is based on a Voxel-Expanded Gathering method to fundamentally reduce the workload of the data structuring step in the inference phase. The initial prototype of HgPCN has been implemented on an Intel PAC (Xeon+FPGA) platform. Four commonly available point cloud datasets were used for comparison, running on three baseline devices: Intel Xeon W-2255, Nvidia Xavier NX Jetson GPU, and Nvidia 4060ti GPU. These point cloud datasets were also run on two existing PCN accelerators for comparison: PointACC and Mesorasi. Our results show that for the inference phase, depending on the dataset size, HgPCN achieves speedup from 1.3× to 10.2× vs. PointACC, 2.2× to 16.5× vs. Mesorasi, and 6.4× to 21× vs. Jetson NX GPU. Along with optimization of the memory-intensive down-sampling bottleneck in pre-processing phase, the overall latency shows that HgPCN can reach the real-time requirement by providing end-to-end service with keeping up with the raw data generation rate.more » « less
-
The proliferation of innovative mobile services such as augmented reality, networked gaming, and autonomous driving has spurred a growing need for low-latency access to computing resources that cannot be met solely by existing centralized cloud systems. Mobile Edge Computing (MEC) is expected to be an effective solution to meet the demand for low-latency services by enabling the execution of computing tasks at the network-periphery, in proximity to end-users. While a number of recent studies have addressed the problem of determining the execution of service tasks and the routing of user requests to corresponding edge servers, the focus has primarily been on the efficient utilization of computing resources, neglecting the fact that non-trivial amounts of data need to be stored to enable service execution, and that many emerging services exhibit asymmetric bandwidth requirements. To fill this gap, we study the joint optimization of service placement and request routing in MEC-enabled multi-cell networks with multidimensional (storage-computation-communication) constraints. We show that this problem generalizes several problems in literature and propose an algorithm that achieves close-to-optimal performance using randomized rounding. Evaluation results demonstrate that our approach can effectively utilize the available resources to maximize the number of requests served by low-latency edge cloud servers.more » « less
-
Users on edge generate deep inference requests continuously over time. Mobile/edge devices located near users can undertake the computation of inference locally for users, e.g., the embedded edge device on an autonomous vehicle. Due to limited computing resources on one mobile/edge device, it may be challenging to process the inference requests from users with high throughput. An attractive solution is to (partially) offload the computation to a remote device in the network. In this paper, we examine the existing inference execution solutions across local and remote devices and propose an adaptive scheduler, a BPS scheduler, for continuous deep inference on collaborative edge intelligence. By leveraging data parallel, neurosurgeon, reinforcement learning techniques, BPS can boost the overall inference performance by up to 8.2× over the baseline schedulers. A lightweight compressor, FF, specialized in compressing intermediate output data for neurosurgeon, is proposed and integrated into the BPS scheduler. FF exploits the operating character of convolutional layers and utilizes efficient approximation algorithms. Compared to existing compression methods, FF achieves up to 86.9% lower accuracy loss and up to 83.6% lower latency overhead.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1145/3579371.3589113
