Researchers propose Joules-to-GDP framework to measure AI energy efficiency
Understand this story
Amirreza Zeinali and other researchers proposed the Joules-to-GDP framework, using two metrics to measure the economic output per unit of energy consumption of AI: Net Value per Joule (NVJ) calculates the output dollar value minus the AI execution cost, then divides by energy consumption; Time Saving per Joule (TSJ) calculates expert time saved divided by energy consumption. The researchers tested the framework on 220 GDPval tasks across 44 occupations. The report did not give the test results.
AI overview based on source reports · Updated 8 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIJoules-to-GDP: How Much Economic Value Do AI Systems Produce per Unit of Energy?
The paper introduces Joules-to-GDP, a framework measuring Net Value per Joule (NVJ) and Time Saving per Joule (TSJ), and applies it to 220 GDPval tasks across 44 occupations. Among five tested models, GLM 5.2 delivers the most net value at $72.27 per task on average, while Gemma 4 31B reaches 56% of that with 23% of the energy, for 2.5x the NVJ and 2.8x the TSJ; eight H100s use 3.2x the energy of one to cut completion time by only 22%. Measured against the work of experts who use AI, GLM 5.2's NVJ falls by 70% and its TSJ by 65%.
Activity for this story
- Current comparable activity
- 8
- Peak comparable activity
- 10Nov 6 06:00
- Change in the past 24 hours
- –
Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.