GRAMARCH: A GPU-ReRAM based Heterogeneous Architecture for Neural Image Segmentation

Joardar, Biresh Kumar; Jayakodi, Nitthilan Kannappan; Doppa, Janardhan Rao; Li, Hai; Pande, Partha Pratim; Chakrabarty, Krishnendu

doi:10.23919/date48585.2020.9116273

Cited by 9 publications

(8 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, 3D NoC enables high-throughput multicast support [12]. As a result, 3D NoC is more suited for long-range, multicast, and many-to-one-to-many traffic patterns [12], [13]. Hence, we conjecture that 3D NoC is a suitable communication backbone to enable the proposed ReRAMbased architecture for training GNNs.…”

Section: Introductionmentioning

confidence: 95%

“…By stacking multiple layers above each other, the physical distance between PE tiles is reduced significantly [13]. In addition, 3D NoCs enable high-performance multicast support [12]; both these characteristics are beneficial for GNN training. In this work, we use a 3D mesh NoC as interconnection backbone because it can support 3D tree multicast.…”

Section: B Role Of 3d Nocmentioning

confidence: 99%

“…In addition, the mapping strategy influences on-chip traffic and it must complement the underlying NoC design to ensure the best performance. In this work, we employ a simple Simulated Annealing (SA)-based strategy for optimal mapping following [12] as it can uncover high-quality solutions in a reasonable amount of time. The mapping of weights and the 𝐴𝑑𝑗 matrix to the PEs can be envisioned as a combinatorial optimization problem: Given a total of 𝑃 PEs and 𝐿 layers (V and E), our aim is to distribute all the computation layers such that the highly communicating layers are mapped to nearby PEs.…”

Section: Overall Mapping To Pesmentioning

confidence: 99%

“…Prior work has shown that 3D NoCs enable the design of a low latency and high throughput communication backbone for manycore chips [11]. In addition, 3D NoC enables high-throughput multicast support [12]. As a result, 3D NoC is more suited for long-range, multicast, and many-to-one-to-many traffic patterns [12], [13].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

ReGraphX: NoC-enabled 3D Heterogeneous ReRAM Architecture for Training Graph Neural Networks

Arka

Doppa

Pande

et al. 2021

2021 Design, Automation &Amp; Test in Europe Conference &Amp; Exhibition (DATE)

Self Cite

View full text Add to dashboard Cite

Graph Neural Network (GNN) is a variant of Deep Neural Networks (DNNs) operating on graphs. However, GNNs are more complex compared to traditional DNNs as they simultaneously exhibit features of both DNN and graph applications. As a result, architectures specifically optimized for either DNNs or graph applications are not suited for GNN training. In this work, we propose a 3D heterogeneous manycore architecture for on-chip GNN training to address this problem. The proposed architecture, ReGraphX, involves heterogeneous ReRAM crossbars to fulfill the disparate requirements of both DNN and graph computations simultaneously. The ReRAMbased architecture is complemented with a multicast-enabled 3D NoC to improve the overall achievable performance. We demonstrate that ReGraphX outperforms conventional GPUs by up to 3.5X (on an average 3X) in terms of execution time, while reducing energy consumption by as much as 11X.

show abstract

Section: Introductionmentioning

confidence: 95%

Section: B Role Of 3d Nocmentioning

confidence: 99%

Section: Overall Mapping To Pesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

ReGraphX: NoC-enabled 3D Heterogeneous ReRAM Architecture for Training Graph Neural Networks

Arka

Doppa

Pande

et al. 2021

2021 Design, Automation &Amp; Test in Europe Conference &Amp; Exhibition (DATE)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Today, many hardware options give the users flexibility when chosen for training and testing. These options include GPUs [3][4], TPUs [5] [6], FPGAs [7] [8], heterogeneous [9] and servers [10].…”

Section: Introductionmentioning

confidence: 99%

U-Net Cost Analysis Using Roofline Model

Salim¹,

Dawwd²,

Ali³

2022

(AREJ)

View full text Add to dashboard Cite

One of the most important challenges facing U-Net architecture performance is the method design of its components and how to choose the suitable hardware computing device to deal with the training labelled datasets. Convolution is the most process that requires computations and memory costs, which is needs to minimize. Thus, one of the suitable selection is to change the type of the convolution. Other suggested solutions are to reduce the size of image, number of bits, and, stride value, in addition to number of filters, and image batches. Therefore, in this paper the roofline model will used as performance guide in analyzing the FLOPs and the memory bandwidth boundaries of a U-Net model with different configurations. The cost has been assessed with compared to the limitation of three computing devices, GPU230MX, GPU940MX and GPU2060rtx super. 128 × 128 image dataset has been used during the U-Net costperformance evaluation process. Based on the analysis, the evaluation results show that the solution that achieves a balance between memory and computations is to implement a U-Net model in parallel using RTX2060 super card with the configurations of batch size is 16, image size of 128×128, number of bits is 32, shared memory management.

show abstract

Scalable and Energy-Efficient NN Acceleration with GPU-ReRAM Architecture

Moura,

Carro

2023

Applied Reconfigurable Computing. Architectures, Tools, and Applications

View full text Add to dashboard Cite

GRAMARCH: A GPU-ReRAM based Heterogeneous Architecture for Neural Image Segmentation

Cited by 9 publications

References 15 publications

ReGraphX: NoC-enabled 3D Heterogeneous ReRAM Architecture for Training Graph Neural Networks

ReGraphX: NoC-enabled 3D Heterogeneous ReRAM Architecture for Training Graph Neural Networks

U-Net Cost Analysis Using Roofline Model

Scalable and Energy-Efficient NN Acceleration with GPU-ReRAM Architecture

Contact Info

Product

Resources

About