Action: OECD AI Compute Survey

[David Kanter]

Hi Everyone,

I hope you are well!

As some of you may know, I'm a co-chair of the OECD Task Force on AI Compute (https://oecd.ai/en/network-of-experts/working-group/1136). You can view the task force’s full membership here (https://oecd.ai/en/network-of-experts/working-group/1136) and read a recent article about our work on the OECD.AI Wonk (http://oecd.ai/).

Basically, we are looking to create easily understandable indices for AI Compute Capability. It's not quite benchmarks, but it is closely related and I believe there is potential overlap with MLPerf.

The task force has launched a survey to gather first-hand information on existing indicators and frameworks to measure AI computing resources (“AI compute”) as well as related environmental impacts. Your input through the survey would be invaluable.

The survey is intended for both users and suppliers of AI compute or those with in-depth technical knowledge on the subject. The survey includes an optional section specifically on the environmental impacts of AI compute, for those with relevant expertise. Please feel free to share it with other relevant colleagues in your network through LinkedIn (https://www.linkedin.com/posts/oecd-ai_aicompute-oecdai-artificialintelligence-activity-6912367267863306240-SSXq?utm_source=linkedin_share&utm_medium=member_desktop_web) and Twitter (https://twitter.com/OECDinnovation/status/1506616430359371789).

The deadline to complete the survey is 6 April 2022. The survey should take between 10 and 15 minutes to complete and can be accessed here: https://survey.oecd.org/index.php?r=survey/index&sid=679658&newtest=Y&lang=en

For further questions, please do not hesitate to contact A...@oecd.org or send me a line with any feedback.


Thanks,

David Kanter

Executive Director

MLCommons