Java Runtime Optimization Enhancements
Budget: €30 – €250 EUR
I’m looking to extend the standard javac pipeline so that selected compile-time transformations flow seamlessly into runtime-level optimizations. The sole objective is measurable performance gains, and the hot spot I need addressed is the handling of function calls.
Here’s the current thinking:
• During compilation I want to inject bytecode-level hints or re-write patterns that the JVM’s JIT can exploit immediately at runtime.
• The focus is on aggressive inlining, de-virtualization, and any other call-site refinement that reduces dispatch overhead without breaking portability.
• I expect clear before-and-after benchmarks (micro-benchmarks are fine) proving the speed-up on a standard HotSpot JVM.
Please bring practical experience with javac plugins, annotation processors, ASM or similar bytecode toolkits, and a solid grasp of JIT heuristics. Deliverables are the source of the transformation module, build instructions, and a short report explaining the rationale and the observed performance lift.
If this sounds within your wheelhouse, let’s dive in.
Here’s the current thinking:
• During compilation I want to inject bytecode-level hints or re-write patterns that the JVM’s JIT can exploit immediately at runtime.
• The focus is on aggressive inlining, de-virtualization, and any other call-site refinement that reduces dispatch overhead without breaking portability.
• I expect clear before-and-after benchmarks (micro-benchmarks are fine) proving the speed-up on a standard HotSpot JVM.
Please bring practical experience with javac plugins, annotation processors, ASM or similar bytecode toolkits, and a solid grasp of JIT heuristics. Deliverables are the source of the transformation module, build instructions, and a short report explaining the rationale and the observed performance lift.
If this sounds within your wheelhouse, let’s dive in.