All Tags
AWS
algorithm-design
architecture
cloud-principles
cost-reduction
data-centric
data-compression
data-processing
deployment
design
edge-computing
energy-footprint
hardware
libraries
locality
machine-learning
management
measured
migration
model-optimization
model-training
performance
queries
rebuilding
scaling
services
strategies
template
workloads
Tactic: Use Computation Partitioning
Tactic sort:
Awesome Tactic
Type: Architectural Tactic
Category: green-ml-enabled-systems
Title
Use Computation Partitioning
Description
Computation partitioning is the process of dividing the computations of a convolutional neural network (CNN) between a mobile client and a cloud server. The goal is to optimize energy consumption and efficiency. The NeuPart framework (Manasi et al 2023) is an example of a partitioning approach. NeuPart divides computational tasks between the mobile device (client) and the remote server or data center (cloud) in real time based on energy consumption. By offloading computationally intensive tasks to the cloud and executing lighter tasks locally, NeuPart resulted in significant energy savings of up to 52% in cloud-based computations.
Participant
Software Designer
Related software artifact
Neural Networks
Context
Cloud
Software feature
< unknown >
Tactic intent
Improve energy efficiency dividing computational tasks between the mobile device (client) and the remote server or data center (cloud) in real-time based on specific conditions or requirements.
Target quality attribute
Energy Efficiency
Other related quality attributes
< unknown >
Measured impact
Manasi et al demonstrated that at a certain effective bit rate and transmission power, the optimal partition for specific CNN models resulted in energy savings of up to 52.4% over a fully cloud-based computation and 27.3% over a fully in situ computation.
Source
Susmita Dey Manasi, Farhana Sharmin Snigdha, and Sachin S Sapatnekar. 2020. Neupart: Using Analytical Models to Drive Energy-Efficient Partitioning of CNN Computations on Cloud-Connected Mobile Clients. IEEE Transactions on Very Large-Scale Integration (VLSI) Systems 28, 8 (2020), 1844–1857. (DOI: https://doi.org/10.1109/TVLSI.2020.2995135)Graphical representation
