Keynote: Finding Your Power to Accelerate to a Sustainable Future
summary
Data growth and AI compute are driving energy demand—AI power doubles roughly every four months and data centers risk using about 8% of global electricity by 2030—while a survey of 100+ companies shows average CPU utilization at only 20–40%, wasting power. Demonstrated tools include a carbon aware scheduler for power-based workload placement, a power manager that provisions CPU power features, and telemetry via collectd and telegraf. Claim: language choices can change energy use up to ~75x for large apps; goal: 10× processor efficiency by 2030 and net zero by 2040.
AI-written summary · about
chapters (5)
related talks
01h 02mJoint Keynote Address: Scaling AI Sustainably: An Uncharted Territory
20mBuilding GenAI That Doesn't Cost the Earth
41mSustainable Digital
12mKeynote: Cloud Native Sustainability - How You Can Help Right Now!
25mBuilding a Sustainable Future
44mFighting Climate Change by Building Sustainable Software
29mHow AI and Deep Learning could effort the Climate Change
38mBetter Coding for the Climate?!