Skip to content

FastBytes 0.1.1 — FastSIMD Hardware Acceleration Engine Integration

Latest

Choose a tag to compare

@andrestubbe andrestubbe released this 13 Aug 11:43
· 4 commits to main since this release

TITLE:
FastBytes: High-Performance SIMD Byte Array Operations for Java 🚀

DESCRIPTION:
Demonstrating FastBytes, the SIMD-powered byte processing engine for Java. Powered by the FastSIMD vector engine, it accelerates byte scanning, memory copying, zero-filling, xxHash32 hashing, and 4K video frame XOR transformations.

TECHNICAL HIGHLIGHTS:

  • AVX2 Vector Engine: Integrated with FastSIMD for 32-byte parallel delimiter matching and 256-bit memory copies.
  • 4K Video Frame Glitch XOR: Real-time bitwise XOR transformation at over 120 FPS.
  • Zero-Allocation Search: Sub-50ms 100MB buffer scanning without GC pause overhead.
  • Official JMH Suite: Standardized benchmark Uber-JAR executable via run-benchmark.bat.

RESULTS:

  • SIMD Byte Search: 15.3x speedup over standard Java loops (38.4 ms vs 589.0 ms)
  • 4K Frame Glitch XOR: 111+ FPS real-time rendering
  • JMH Throughput: 128,000+ ops/s

FASTJAVA CORE REPOSITORIES:

  1. FastBytes: https://github.com/andrestubbe/FastBytes
  2. FastSIMD: https://github.com/andrestubbe/FastSIMD
  3. FastMemory: https://github.com/andrestubbe/FastMemory
  4. FastPointer: https://github.com/andrestubbe/FastPointer

Ecosystem Overview: https://github.com/andrestubbe/FastJava

Part of the FastJava Ecosystem — Making the JVM faster.
Made with ⚡ by Andre Stubbe

TAGS:
Java, FastJava, FastBytes, SIMD, AVX2, Byte Array, High Performance, C++, JNI, JMH Benchmark