04-27, 14:45–15:30 (America/Los_Angeles), Kodiak Theatre
As the impact of climate change has gradually presented itself in our daily lives, we have to take
actions to mitigate its effects. United Nations SDGs goal is to reach net-zero carbon
dioxide(CO2) emissions by 2050. To meet this goal, we can start to reduce CO2 emission from
daily programming and computing usage. Have you understand the amount of CO2 emission
from a Pytorch-based deep learning model? Do you know how to choose the optimal hardware
and cloud computing resources to reduce training time and energy, in order to eliminate CO2
emission? This talk will share the state of art calculator software and cloud usage approaches
via different regions and time scales to save our planet.
This talk is targeted at the beginner to intermediate developers. I’ve immersed myself with the concept and tools for green programming and sustainable computing. In this talk, I will share my learning journey and a short demo to inform people how much CO2 emission is produced by our daily computer services. In general, there are two types of approaches to tackle with estimating CO2 emission from model programs and cloud computing resources.
1. Programming codes: use the “code carbon” and “carbon aware SDK” open source software to estimate the CO2 emission from training a deep learning model. This approach can help developers to optimal runtime for code chunks in order to reduce CO2 emission.
2. Cloud computing: use the “cloud carbon footprint” open source software to compare the energy use and carbon emissions (metric tons CO2e) from public cloud usage (AWS, GCP, Azure). This approach can help developers to select optimal hardware and a data center region globally with efficient renewable energy usage and minimize the CO2 emission.
The ultimate goal of this talk is to help developers to be aware of carbon footprint while working on model programming and cloud computing usage. By doing this, we can reduce the increasing risks from climate change in our daily lives.
No previous knowledge expected
Ying-Jung is an environmental scientist turned data scientist / machine learning engineer. She received her Ph.D. in environmental science and management school at UC Santa Barbara. Also, she had been a post-doc at eScience institute at University of Washington. She started her data science/ machine learning career in the greater Seattle area. She is extremely interested in applying models with different domain business problems such as, finance, agriculture, and hydroclimate. She is also keen on learning an efficient approach to use public cloud computing services. Also, she had learned the state of art machine learning algorithms via reading papers, attending online/in-person meetups, and joining major machine learning conferences. Finally, she is a PyData impact scholar in 2021. She'd like to share her learning journey in data science and sustainability with people. In her free time, she likes to go hiking and play basketball / pickleball.