2024.04.23

[seanmcilroy29]

---

2024.04.23 Agenda/Minutes

---

Time 0900 (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

Discussion

• Define the process for converting the spreadsheet data into an impact framework data source - Joseph

Renewable Energy Percentage

• Completed defining language for the column headers
  Action Item: Members requested to assist in defining column headers GSF Real-Time Cloud Renewable Energy Percentage spreadsheet
• RTC Cloud Region Impact Framework Documentation

Discussion Board

• Review miro dataflow in detail
  • Continue to review and approve the proposed Scope.

Open Issues Project Board

For Review

• How should GPU energy use be estimated? #37

Open Issues for review

• Fill out geolocation column for Cloud Region impact framework #31
• Fill out Watttime column for Cloud Region impact framework #32
• Fill out Electricity Maps data for Cloud Region impact framework dataset #33
• Decide on source and fill in energy production data for Cloud Region impact framework dataset #34
• Complete CFE region and any other data gaps in Cloud Region dataset #35
• AWS PUE Data Alignment #29
• GCP PUE Data Alignment #28
• Azure PUE data alignment #27
• Cloud provider Power Purchase Agreement (PPA) generation locations as an Impact Framework dataset #20
• Windows Energy Estimation Tool #19
• What are the use cases for real time power and carbon intensity information? #12
• How can energy data collection and reporting be implemented by a cloud provider? #4

AOB

• Topics added during the meeting

Previous Meeting Action Items

• Follow up with Thomas from GCP to get an update on collecting Azure PUE data methodology.
• Follow up with Chris Walker from AWS to get an update on any new PUE data that may be available.
• Create an issue to define the process for converting the spreadsheet data into an impact framework data source. Assign the issue to Joseph.
• Research GPU power metrics and interfaces and find an expert to help with collecting GPU power data.
• Reach out to contacts at NVIDIA to explore sharing GPU power curves and metrics.
• Map the Google annual environmental report region names to the actual cloud region names and clean up the data.
• Invite Marco Stiller and/or Chris from the Green Software Foundation to the next meeting to discuss their carbon footprint tool.

AOB

• Any topics members would like to submit

Next Meeting

• 07th May

Adjourn

• Motion to adjourn

---

Members to review and comment on the RTC Cloud Region Impact Framework Documentation
Add a column for Google's 24/7 carbon intensity data once it is approved for publishing.
Members requested to assist in defining column headers GSF Real-Time Cloud Renewable Energy Percentage spreadsheet
Fill in missing data columns in the spreadsheet, including PUE data and IEA annual average numbers.
Get the spreadsheet definitions finalized to start filling in the data. Clarify column names and descriptions.
Review and comment on Scope Flow

[seanmcilroy29]

Attended

[seanmcilroy29]

Cloud meta data plugin - https://github.com/Green-Software-Foundation/if-plugins/blob/main/src/lib/cloud-metadata/gsf-data.csv

[PindyBhullar]

Attended from 4.30pm

[adrianco]

Attended

[seanmcilroy29]

MoM
Adrian opens the meeting at 0900 (PT) / 1600 (BST)

Cloud-based carbon footprint estimation and data sources.
Joseph and Adrian had a conversation about a plug-in for a data analysis tool that uses a smaller version of a data set. They went through a file that had documentation on the plug-in's fields and background information on where the information comes from. Joseph shared a link to a readme file for a cloud-based state plugin. Adrian explained how their company uses third-party APIs to gather carbon intensity data for various regions. They discussed an earlier version of the dataset and agreed to update their impact framework to reflect the latest version.
Integrating cloud metadata plugin with impact framework.
Joseph explains how to use the cloud metadata plugin for carbon intensity reporting. Joseph and Adrian discuss updating a plugin to expose new data. Adrian and Joseph discuss whether to include region data in the cloud metadata plugin.

Cloud metadata impact framework implementation.
Adrian and Joseph discussed how to manage missing data in a plugin. Adrian proposed raising an exception and dumping it into the console, and he emphasized the importance of PUE (Power Usage Effectiveness) while performing calculations, as it is a useful number. They also discussed how to handle missing data in their plugin. Adrian suggested using the higher end of a range provided by AWS as a default value. On the other hand, Joseph recommended returning a panic if a requested value is missing, but Adrian preferred using annual average data as a starting point for those who don't have API access. Later, Adrian and Vincent talked about adding metadata to an impact framework for cloud instances.
In cloud computing, PUE estimates, CPU time disclosure, and GPU energy estimation.
Adrian and Vincent discussed the estimated PUEs for different GCP regions. They also discussed the lack of CPU time disclosure for AWS and Azure instances and the absence of GPU energy estimation plugins.

GPU power usage and data sources for real-time monitoring.
Adrian has discovered a data source that provides real-time information on GPU power usage, which can be instrumental in monitoring and managing power consumption in data centers. GPUs consume a lot more power than CPUs, making it a critical issue in the industry. Adrian and Vincent discuss the challenges associated with accurately estimating GPU power usage, and they mention the DMTF Redfish interface as a standard method for collecting power usage data in datacenter environments.

Energy monitoring tools for CPUs and GPUs.
Adrian is planning to finalize the update for the impact framework and make it available for public use. Sean will get in touch with Adrian to discuss the initial impact framework and schedule a podcast discussion. During a conversation between Adrian and Pindy, they talked about energy monitoring agents for cloud instances and Scott Bender was mentioned as the only other monitoring agent available. Pindy and Adrian also discussed various tools for monitoring and managing computer systems, including Skyfinder. Skyfinder is a French diving helmet that comes with an IPMI interface which provides power information.

Using impact framework to measure cloud providers' carbon footprint.
Adrian and the team had a discussion about reaching out to a Bovitza to attend their meetings. Adrian talked about the importance of reaching a compromise with the organisation and updating their impact framework with carbon data. The team also discussed the potential benefits of focusing on real-time data sources, like Prometheus, to improve the performance of their cloud provider. They also talked about some potential requests they could make to their cloud providers, such as asking for better TDP numbers and real-time instance data.

Action Items

• Update the cloud metadata plugin readme to include information on the regional metadata dataset.
• Provide a pull request updating the GSF data CSV file with the current regional metadata dataset.
• Contact Hubblo to discuss collaborating on energy monitoring and potentially joining a future meeting.
• Investigate available power/energy metrics for AWS training instances and report findings.
• Check in with the Cloud Carbon Footprint initiative to explore using regional metadata datasets to improve their data.

[seanmcilroy29]

Approved