2024.07.30

[seanmcilroy29]

---

2024.07.30 Agenda/Minutes

---

Time 0800 (PT) / 1600 (BST) - See the time in your timezone

• Chair – Adrian Cockcroft
• Chair - Pindy Bhullar (UBS)
• Convener – Sean Mcilroy (Linux Foundation)

Antitrust Policy

Joint Development Foundation meetings may involve participation by industry competitors, and the Joint Development Foundation intends to conduct all of its activities in accordance with applicable antitrust and competition laws. It is, therefore, extremely important that attendees adhere to meeting agendas and be aware of and not participate in any activities that are prohibited under applicable US state, federal or foreign antitrust and competition laws.

If you have questions about these matters, please contact your company counsel or counsel to the Joint Development Foundation, DLA Piper.

Recordings

WG agreed to record all Meetings. This meeting recording will be available until the next scheduled meeting.

Roll Call

Please add 'Attended' to this issue during the meeting to denote attendance.

Any untracked attendees will be added by the GSF team below:

• Full Name, Affiliation, (optional) GitHub username

Agenda

• Approve agenda
• Approve previous Meeting Minutes

Review

• [Documentation] Artificial Intelligence Act #54
• Review the documentation files on GitHub and provide updates.
• https://github.com/Green-Software-Foundation/real-time-cloud/blob/Dev/Cloud_Region_Metadata_specification.md
• https://github.com/Green-Software-Foundation/real-time-cloud/blob/Dev/AWS_renewable_energy_2023.csv

Discussions

• GCP Water Usage Effectiveness alignment #47
• AWS Water Usage Effectiveness alignment #48
• Updated GCP data for 2023 #51
• Updated AWS data for 2023 #52
• How should GPU energy use be estimated? #37

Ref Material

• GSF Real-Time Cloud Renewable Energy Percentage spreadsheet
• RTC Cloud Region Impact Framework Documentation
• miro dataflow in detail
  • Continue to review and approve the proposed Scope.

Next Steps

• Is the data set ready for release

AOB

• Topics added during the meeting

Next Meeting

• 13 Aug

---

[adrianco]

attended

[kapokasa]

attended

[seanmcilroy29]

MoM

Cloud region metadata and sustainability.
Chris Xie and Adrian discuss a pull request in GitHub for a project they are working on. Adrian explains how to analyze cloud provider data for regions, providers, and years.

Cloud providers' carbon footprint, AWS's green claims, and standards for measuring efficiency.
Adrian recently wrote an article on Medium where he compared the annual reports of AWS, Google, and Microsoft. The article received a good amount of traffic. In a recent meeting, the team discussed using a cloud carbon footprinting tool on GCP to gather more specific emissions data. Chris Xie and Adrian are considering publishing data to a standards group and Adrian plans to attend a meeting to gather feedback. During the meeting, Adrian expressed concerns about AWS's use of the ISO 16000-22 (Sei) standard, pointing out a lack of transparency and accuracy in their calculations. An Accenture representative mentioned the Accenture AWS piece, which discusses their approach to Sei calculation.

Cloud computing carbon footprint and energy usage.
Adrian is working on documenting and organizing tools for measuring carbon emissions in AI systems. Adrian mentioned that AWS generates 100% green energy in 22 regions and sells excess energy to other areas. Amazon has added Japan to the list of regions with 100% energy coverage and increased the deployment of energy PPAs in Japan. Adrian and Chris discussed the importance of keeping energy data up to date, with Adrian emphasizing the need for real-time data. Chris expressed interest in joining the conversation later. Adrian updated their collected data, including information on Microsoft's 2023 report and the need for more accurate and up-to-date data.

Cloud carbon accounting methodology and industry standards.
The group discussed the detailed methodology for multi-tenant carbon accounting in the cloud, which could be useful for others. The proposed approach involves internal accounting groups and revenue-based allocation without explicitly mentioning AI. Pindy asks about similar methodologies from AWS and Microsoft. Adrian replies that Accenture is the nearest thing to AWS, and Microsoft has a detailed methodology paper, but not in this much detail. Adrian notes that Google is doing better on transparency with actual data and more fine-grained information, and the group agrees that they need to try to get everyone to provide the same data. Adrian and Henry discuss the Sei standard and its control, with Adrian expressing frustration at the lack of control. Adrian hopes Accenture will explain their actions and plans to chat with Henry and others on Wednesday and Thursday.

Action Items

• Review the cloud region metadata specification document and pull request.
• Read the new Google paper on multi-tenant carbon accounting methodology.
• Schedule a meeting with Microsoft for updated data centre efficiency figures.
• Contribute any missing/incorrect data to the cloud region metadata spreadsheet.

[PindyBhullar]

Attended