2024.05.23

[seanmcilroy29]

---

2024.05.23 Agenda/Minutes

---

Time 1600 (GMT) - See the time in your timezone

• Co-Chair - Henry Richards (WattTime)
• Co-Chair - Naveen Balani - (Accenture)
• Co-Chair – Abhishek Gupta (BCG)
• 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

Use Case Review

• Eco-CI
• Executive Summary - a free plugin called Eco-CI that leverages an open source Machine Learning Model implementation called Cloud Energy to estimate the Energy and CO2 consumption as SCI of a CI / CD pipeline run on major platforms like Github and Gitlab

Techstrong TV Sustainability Interview with Bonnie Schneider - Sean

• Techstrong Interview with SWG member(s) to discuss the SCI
• Volunteer names
  • Chris Xie (Future Wei)
  • Navveen Balani (Accenture)
  • Greg Rivera (CAST)
  • Tammy McClellan (Microsoft)

GreenAI - Discussion Continued - Henry / Thomas

• Green AI Committee update/rollout plan
  • Green AI Project Proposal Early Brief.pdf

What next for the SCI

• Are there any areas of the SCI that we should expand on or include?
• [Discussion] SCI aggregation views and intentions #111

Project Review updates

• Real-Time Cloud - Adrian
  • Defining the GSF Real-Time Cloud Renewable Energy Percentage spreadsheet.
  • collaboration documentt to create the initial header explanations.
• Software Carbon Efficiency Rating - Review - Chris Xie
  • Discussion/review focuses on the LLM and proposed methodology for LLM in generative AI.
• Transforming Organisational Software Sustainability
• sci-guide - Navveen

Policy WG Request

• GSF Policy Radar

Articles

• Nothing to review

For Review

• Lifecycle Assessment (LCA) for Software Applications using SCI and Impact Framework - Srini

Note:
WG use case template submission - After submitting this issue, your use case will be submitted to the WG Agenda for discussion.
Article submission—Once you submit this issue, it will be assigned to the GSF Editor for review.

Future meeting Agenda submissions

• If you have any Agenda submissions, please let Sean know via email at [email protected] or open an Issue

Next Meeting

• 06th Jun

Adjourn

• Motion to adjourn

---

Standing Agenda / Future Agenda submissions

• Standing Agenda Item: The First meeting of each month is a member Introduction / Buddy-buddy meeting to help new members understand the work in the SWG.
• Standing Agenda Item - First meeting of each month - 5 - 10 min round up of what is happening across the other WG

[marcoshidalgonunes-avanade]

Attended

[Henry-WattTime]

Attended

[ArneTR]

attended

[seanmcilroy29]

Attended

[sk16-dev]

Attended

[tmcclell]

Attended

[kapokasa]

Attended

[pilarbilbao]

Attended

[navveenb]

Attended

[seanmcilroy29]

MoM
Henry opens the meeting at 1600 BST

Eco CI - Using machine learning to estimate the energy consumption of GitHub and GitLab pipelines.
Arnie demonstrates the Eco CI by explaining the purpose of the pull request, which includes solutions for reducing energy consumption and open-source tools for improving efficiency. Arnie demonstrates the Eco CI plugin, which provides data on energy and carbon usage for CI/CD pipelines. Arnie uses a machine learning model to estimate power consumption in GitHub repositories. The open-source, Python-based model is integrated into GitHub and GitLab pipelines. Arnie explains how their tool can optimise pipeline costs by analysing energy consumption data. They demonstrate the tool's functionality by showing a screenshot of the output.

Using machine learning to predict CPU performance based on spec power data.
Asim inquires about the underlying values embodied in the SCI score. Arnie explains that they are unaware of the underlying mechanism, which would be a constant for the specific case. Arni and Asim discuss using spec power to construct a power curve based on actual hardware. Arni explains that they employ a Python machine learning model to train on 1000 data lines to obtain a nonlinear curve. Asim and Sean discuss the definition of "run" in the context of a machine-learning model.

Integrating energy measurement tool with impact framework.
Asim suggests integrating energy measurement into the impact framework, CI runs. Arni discusses their tool, which provides green metrics for open-source projects, focusing on GitHub shared runners. Arni highlights the challenges of detecting code changes in oversubscribed environments and the potential for optimization based on variability analysis.

SCI equation implementation and adoption in green ICT projects.
Asim and Henry discuss the importance of providing detailed information on how SEO calculations were implemented in case studies. They emphasize the value of showcasing practical examples of SEO implementation to inspire adoption and encourage further development. SC organizations were represented in a panel discussion about SCI development. Thomas will lead the GreenAI committee.

Software carbon footprint and emissions reporting.
Asim explains that KPIs are not meant to be applied to everything, including air travel, and there are different types of KPIs for different purposes. Henry, Tammy, and Asim discuss the use of KPIs in software development and the importance of understanding their limitations and context. Asim discusses the importance of considering the entire organization rather than just one product or department when addressing green software. Sean O adds that optimising one aspect of software can come at the expense of others and that a holistic approach is necessary to ensure sustainability. Sean O and Asim discuss the scope of SCI in terms of granularity and emissions reduction. Asim suggests expanding the scope to include carbon purple King metric and inclusions for the next version.

Maintaining and organizing datasets within the organization.
Henry suggests focusing on VSP er, while Asim and Sean discuss comparability tools.
Henry and Sean discuss the evolution of real-time cloud data and the potential repository location. Asim questions whether to centralise or decentralise data in separate repos, with pros and cons to each approach. Asim suggests using a single data repository to make data easier to maintain and access. Henry and Sean discuss the need for a single point of contact for data requests and updates.

Software sustainability standards and their potential adoption in a regulatory perspective.
Asim highlights the importance of concrete evidence for USCI score, citing an FT article on greenwashing and small nonprofits' ratings. Chris outlines that adopting standards generated by the Green Software Foundation can lead to greater regulatory adoption. Asim suggests reviewing SC before consistency review to get feedback early.

Techstrong interview
For the Techstrong interview, the SWG has agreed to the following volunteers to represent them;

• Chris Xie (Future Wei) - [email protected]
• Navveen Balani (Accenture) - [email protected]
• Greg Rivera (CAST) - [email protected]
• Tammy McClellan (Microsoft) - [email protected]

Action Items

• Present SCER to SC at the next meeting for early feedback
• Schedule Techstrong interview for SWG representatives
• Investigate data sources for real-time cloud dataset
• Schedule deeper discussion on SCER specification at the next meeting