Optimizing code saves energy. But does #AI-powered optimization save more than it burns? This #Java experiment puts #GenAI, #JFR, and CodeCarbon to the test. @obierlaire delivers hard numbers.

Decide if it’s worth it: https://javapro.io/2026/02/13/can-genai-help-reduce-energy-used-by-java-code/

#Performance #AIAgent @anthropicai

Still treating immutability as a startup-time penalty? #Java25’s Stable Values change that contract. @hannotify dives into the feature that every backend dev should understand early.

Learn it before it’s default: https://javapro.io/2026/02/12/heres-java-25-ready-to-perform-to-the-limit/

#Java #JVM #JFR @openjdk @Oracle

The #JVM warmup tax is real & #Java25 finally attacks it head-on. AOT profiling, Project Leyden speedups, and sharper #JFR events. @hannotify shows where you win #Performance back.

Steal the exact JEPs to cut startup + warmup time: https://javapro.io/2026/02/12/heres-java-25-ready-to-perform-to-the-limit/

#Java @openjdk @Oracle

Как я дал ИИ-ассистенту доступ к JFR, jcmd и jps через MCP

Приложение тормозит. Где узкое место? Почему растёт память? Есть ли дедлок? Я написал JavaPerf — MCP-сервер, который даёт ИИ-инструментам прямой доступ к нативным утилитам JDK для профилирования. Без графических интерфейсов, без ручного CLI — только диалог с ассистентом.

https://habr.com/ru/articles/1007642/

#JavaPerf #MCP #JFR #jcmd #jps #профилирование #производительность #диагностика #AI #трассировка

Как я дал ИИ-ассистенту доступ к JFR, jcmd и jps через MCP

Приложение тормозит. Где узкое место? Почему растёт память? Есть ли дедлок? (Спойлер: обычно всё сразу, и обычно в пятницу после 17:00.) Каждый Java-разработчик хотя бы раз сталкивался с...

Хабр
https://jbachorik.github.io/posts/jfr-mcp-serve.
That's interesting, an MCP server for Java flight recorder data. This allows an agent to optimise Java code automatically.
#java #mcp #jfr #ai #agent
Let the AI Debug It: JFR Analysis Over MCP

Mostly Java/JVM stuff. Focused on serviceability, observability and performance. All opinions are my own and I am not selling any of them.

With Enthusiasm
the devex for grafana #pyroscope is so good.
Im just missing something for heap dump, thread dump and stack trace analysis that lives in
#grafana. the friction of manually grabbing a heap dump/#jfr kills it for me ​
Let the AI Debug It: JFR Analysis Over MCP https://jbachorik.github.io/posts/jfr-mcp-serve
#java #jfr
Let the AI Debug It: JFR Analysis Over MCP

Mostly Java/JVM stuff. Focused on serviceability, observability and performance. All opinions are my own and I am not selling any of them.

With Enthusiasm

Optimizing code saves energy. But does #AI-powered optimization save more than it burns? This #Java experiment puts #GenAI, #JFR, and CodeCarbon to the test. @obierlaire delivers hard numbers.

Decide if it’s worth it: https://javapro.io/2026/02/13/can-genai-help-reduce-energy-used-by-java-code/

#Performance #AIAgent @AnthropicAI

Still treating immutability as a startup-time penalty? #Java25’s Stable Values change that contract. @hannotify dives into the feature that every backend dev should understand early.

Learn it before it’s default: https://javapro.io/2026/02/12/heres-java-25-ready-to-perform-to-the-limit/

#Java #JVM #JFR @openjdk @Oracle

The #JVM warmup tax is real & #Java25 finally attacks it head-on. AOT profiling, Project Leyden speedups, and sharper #JFR events. @hannotify shows where you win #Performance back.

Steal the exact JEPs to cut startup + warmup time: https://javapro.io/2026/02/12/heres-java-25-ready-to-perform-to-the-limit/

#Java @openjdk @Oracle