Search results for: Bayesian Kernel

Authors: Qiuyu Dai, Haochong Zhang, Xiangrong Liu

Abstract:

Efficient matrix-vector multiplication with diagonal sparse matrices is pivotal in a multitude of computational domains, ranging from scientific simulations to machine learning workloads. When encoded in the conventional Diagonal (DIA) format, these matrices often induce computational overheads due to extensive zero-padding and non-linear memory accesses, which can hamper the computational throughput, and elevate the usage of precious compute and memory resources beyond necessity. The ’DIA-Adaptive’ approach, a methodological enhancement introduced in this paper, confronts these challenges head-on by leveraging the advanced parallel instruction sets embedded within Machine Learning Units (MLUs). This research presents a thorough analysis of the DIA-Adaptive scheme’s efficacy in optimizing Sparse Matrix-Vector Multiplication (SpMV) operations. The scope of the evaluation extends to a variety of hardware architectures, examining the repercussions of distinct thread allocation strategies and cluster configurations across multiple storage formats. A dedicated computational kernel, intrinsic to the DIA-Adaptive approach, has been meticulously developed to synchronize with the nuanced performance characteristics of MLUs. Empirical results, derived from rigorous experimentation, reveal that the DIA-Adaptive methodology not only diminishes the performance bottlenecks associated with the DIA format but also exhibits pronounced enhancements in execution speed and resource utilization. The analysis delineates a marked improvement in parallelism, showcasing the DIA-Adaptive scheme’s ability to adeptly manage the interplay between storage formats, hardware capabilities, and algorithmic design. The findings suggest that this approach could set a precedent for accelerating SpMV tasks, thereby contributing significantly to the broader domain of high-performance computing and data-intensive applications.

Keywords: Adaptive method, DIA, diagonal sparse matrices, MLU, sparse matrix-vector multiplication.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 132

Authors: Jaeyoung Lee

Abstract:

Autonomous driving systems require high reliability to provide people with a safe and comfortable driving experience. However, despite the development of a number of vehicle sensors, it is difficult to always provide high perceived performance in driving environments that vary from time to season. The image segmentation method using deep learning, which has recently evolved rapidly, provides high recognition performance in various road environments stably. However, since the system controls a vehicle in real time, a highly complex deep learning network cannot be used due to time and memory constraints. Moreover, efficient networks are optimized for GPU environments, which degrade performance in embedded processor environments equipped simple hardware accelerators. In this paper, a semantic segmentation network, matrix multiplication accelerator network (MMANet), optimized for matrix multiplication accelerator (MMA) on Texas instrument digital signal processors (TI DSP) is proposed to improve the recognition performance of autonomous driving system. The proposed method is designed to maximize the number of layers that can be performed in a limited time to provide reliable driving environment information in real time. First, the number of channels in the activation map is fixed to fit the structure of MMA. By increasing the number of parallel branches, the lack of information caused by fixing the number of channels is resolved. Second, an efficient convolution is selected depending on the size of the activation. Since MMA is a fixed, it may be more efficient for normal convolution than depthwise separable convolution depending on memory access overhead. Thus, a convolution type is decided according to output stride to increase network depth. In addition, memory access time is minimized by processing operations only in L3 cache. Lastly, reliable contexts are extracted using the extended atrous spatial pyramid pooling (ASPP). The suggested method gets stable features from an extended path by increasing the kernel size and accessing consecutive data. In addition, it consists of two ASPPs to obtain high quality contexts using the restored shape without global average pooling paths since the layer uses MMA as a simple adder. To verify the proposed method, an experiment is conducted using perfsim, a timing simulator, and the Cityscapes validation sets. The proposed network can process an image with 640 x 480 resolution for 6.67 ms, so six cameras can be used to identify the surroundings of the vehicle as 20 frame per second (FPS). In addition, it achieves 73.1% mean intersection over union (mIoU) which is the highest recognition rate among embedded networks on the Cityscapes validation set.

Keywords: Edge network, embedded network, MMA, matrix multiplication accelerator and semantic segmentation network.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 410