The Efficiency Paradox: How to Save Yourself & the World
summary
Claims concrete wins: automated LightSwitchOps (a school shell script saves $12,000, AWS shutdown automation cuts cloud bills ~30%) and faster code reduce energy; Quarkus build-time init moves reflection to build step, improving startup, throughput and measured energy use (Quarkus sits below other frameworks on energy chart; native sometimes uses more due to throughput tradeoff). Inventory studies find 12–18% average server utilization, ~25% forgotten servers, ~68% of stored data never read. Queuing math (80%→90% doubles wait) and brain default-mode evidence (72% report best ideas in the shower) justify intentional slack.
AI-written summary · about
chapters (21)
- 0:00Intro
- 2:47Waste
- 5:46e-waste is killing the planet
- 7:20Zombie servers
- 8:35Data
- 9:33Solution: LightSwitchOps
- 13:38Solution: Faster code
- 17:46Case study: Quarkus
- 18:49Java dynamism
- 21:59How do we fix all this?
- 25:26Example: Logging
- 26:04Example: Hibernate
- 26:26Developer joy
- 28:59We have a solution
- 30:52How efficient can we get?
- 33:36Speed vs Resiliency
- 38:15The default mode network
- 42:13Inverse Jevon's paradox
- 44:35Hygge developers
- 45:34Conclusions
- 46:12Outro
related talks
46mWriting Greener Java Applications
49mThe Future is 💚: A Journey through the Energy Efficient ..
50mEnergy Efficient Design Principles of Software Engineering
42mLess waste, more joy, and a lot more green: How Quarkus makes Java better
44mFighting Climate Change by Building Sustainable Software
16mGreen Code: Transforming DevOps for a Sustainable Future
39mCreate an Eco-Friendly World with Green Software Engineering
10mEnergy-Efficient and Cost-Optimized Serverless Computing