Uploaded July 2026 | Updated November 2026, 2 days ago
While Java programs are deployed as bytecode, performance-relevant code will be compiled, at run time, to machine code whose performance rivals that of any other platform. In the HotSpot JVM, this is the job of the just-in-time compilers C1 and C2 and between speculation and traps, inlining and constant folding, escape analysis and scalar replacement (and many more), they apply a number of heuristics, analyses, and transformations to achieve that. This requires expertise, careful research, thorough testing, and regular reevaluations of old assumptions.
In this "Ask the Architects" episode of the Inside Java Podcast, recorded during JavaOne 2026, Nicolai Parlog talks to Roberto Castañeda Lozano, compiler engineer at the Java Platform Group at Oracle.
"Ask the Architects" is our podcast show, where we talk to experts in OpenJDK about their work on the Java language, API, and runtime.
Make sure to check https://inside.java/podcast for more episodes and the youtube.com/playlist?list=PLX8CzqL3ArzUMVSzm-z_-if8BIB55EGl4 playlist.
~~~ Chapters ~~~
0:00 Intro
0:49 Big Picture
4:56 Tiered Compilation
10:11 Speculation & Traps
14:28 Deoptimization
16:42 Hoisting & Sinking
17:39 Heuristics
20:56 Inlining
24:21 Code Quality
27:48 Constant Folding
31:42 Escape Analysis & Scalar Replacement
36:10 Value Types
38:45 Outside Contribution
40:26 Intrinsics
42:48 Working on Compilers
Tags: #Java #OpenJDK #Performance
While Java programs are deployed as bytecode, performance-relevant code will be compiled, at run time, to machine code whose performance rivals that of any other platform. In the HotSpot JVM, this is the job of the just-in-time compilers C1 and C2 and between speculation and traps, inlining and constant folding, escape analysis and scalar replacement (and many more), they apply a number of heuristics, analyses, and transformations to achieve that. This requires expertise, careful research, thorough testing, and regular reevaluations of old assumptions.
In this "Ask the Architects" episode of the Inside Java Podcast, recorded during JavaOne 2026, Nicolai Parlog talks to Roberto Castañeda Lozano, compiler engineer at the Java Platform Group at Oracle.
"Ask the Architects" is our podcast show, where we talk to experts in OpenJDK about their work on the Java language, API, and runtime.
Make sure to check https://inside.java/podcast for more episodes and the youtube.com/playlist?list=PLX8CzqL3ArzUMVSzm-z_-if8BIB55EGl4 playlist.
~~~ Chapters ~~~
0:00 Intro
0:49 Big Picture
4:56 Tiered Compilation
10:11 Speculation & Traps
14:28 Deoptimization
16:42 Hoisting & Sinking
17:39 Heuristics
20:56 Inlining
24:21 Code Quality
27:48 Constant Folding
31:42 Escape Analysis & Scalar Replacement
36:10 Value Types
38:45 Outside Contribution
40:26 Intrinsics
42:48 Working on Compilers
Tags: #Java #OpenJDK #Performance




![Scripting JS and Python with Project Detroit [AtA] Inside Java Podcast 61
OpenJDK recently resurrected Project Detroit, an effort to ease Javas interoperability with Python and JavaScript. Interestingly, while both integrations will work through the _java.scripting_ API, there are still differences between them. On top of that, the JavaScript integration needs to explain how it related to the removal of Nashorn in JDK 16.
In this Ask the Architect episode of the Inside Java Podcast, recorded during JavaOne 2026, Nicolai Parlog talks to Mikael Vidsted, lead of the Java Virtual Machine team at Oracle, about Project Detroit.
Project Detroit: https://openjdk.org/projects/detroit/
~~~ Chapters ~~~
0:00 Intro
0:45 Project Origins & Goals
2:43 JavaScript Integration (Nashorn?)
7:55 Python Integration
10:00 Engine (Version) Selection
13:11 Project Timeline
Make sure to check https://inside.java/podcast for more episodes and the https://www.youtube.com/playlist?list=PLX8CzqL3ArzUMVSzm-z if8BIB55EGl4 playlist.
Tags: #Java #OpenJDK Scripting JS and Python with Project Detroit [AtA] Inside Java Podcast 61](https://i.ytimg.com/vi/pg4NXE9tkIc/mqdefault.jpg)





