All Tags
AWS
ai
algorithm-design
architecture
browser
cloud
cloud-efficiency
cloud-principles
cost-reduction
data-centric
data-compression
data-processing
deployment
design
documentation
edge-computing
email-sharing
energy-efficiency
energy-footprint
enterprise-optimization
green-ai
hardware
libraries
llm
locality
machine-learning
maintainability
management
measured
microservices
migration
mobile
model-optimization
model-training
multi-objective
network-traffic
parameter-tuning
performance
queries
rebuilding
scaling
services
storage-optimization
strategies
tabs
template
testing
workloads
Tactic: Energy Efficient Hardware
Tactic sort:
Awesome Tactic
Type: Architectural Tactic
Category: resource-adaptation
Title
Energy Efficient Hardware
Description
As a practitioner, using hardware designed to closer align with your machine learning workload can reduce energy consumption and cost. Understanding which available hardware is best for your workload and which is more efficient that the other can help reduce energy use and cost. Apply this tactic by comparing available hardware on cloud platforms. As a cloud platform provider, investments in new efficient hardware can be made. Additionally, clear communication or detection for best hardware selection for users.
Participant
Cloud Platform Providers. Machine Learning Practitioners.
Related software artifact
Public Cloud Platforms.
Context
Machine learning systems energy use. Green AI.
Software feature
Design and/or use energy efficient hardware.
Tactic intent
Increased Energy Efficiency.
Target quality attribute
Energy Efficiency.
Other related quality attributes
Cost Optimization.
Measured impact
Example of using Energy Efficient Hardware: Trainium is tailored for training deep learning models, while Inferentia is optimized for inference tasks. Both chips aim to deliver high performance per watt, contributing to energy-efficient AI workloads. Trn1 instances powered by Trainium are up to 25% more energy efficient for DL training than comparable accelerated computing EC2 instances, and Inf2 instances offer up to 50% better performance per watt over comparable Amazon EC2 instances because they and the underlying Inferentia2 accelerators are purpose-built to run deep learning models at scale
Source
Green AI in the Cloud: Energy-Efficient Architectural Tactics for ML-Enabled Systems on Public Platforms by Jingzhi Zhang (Eloise)Graphical representation
