---

📱 Screenshots

---

🚀 Features

• Comprehensive Multi-Component Benchmarking: Tests and evaluates CPU, AI/ML, GPU, RAM, Storage, and Productivity performance in a single platform.
• Advanced Benchmark Modes:
  • Full Benchmark: Runs all 46 tests sequentially to generate overall and category-specific scores.
  • Throttle Test (Stress Test): Sustained heavy CPU + GPU workload (10/30/60 mins) tracking clock speeds, thermal throttling, and performance retention over time.
  • Efficiency Test: Measures performance output relative to power consumption (points/watt) and battery drain.
• Robust Scoring System: Normalized scores relative to a reference baseline (Snapdragon 8 Gen 3) using geometric means to ensure fair evaluation.
• Results History: Local SQLite storage for historical runs with detailed, interactive performance graphs.
• Global Rankings: Compare your device against other benchmarked Android devices with offline/online sync.
• Export & Share: Shareable result cards and raw exports in JSON, CSV, or text formats.
• Premium UI/UX: Fluid design using Jetpack Compose with Material Design 3, real-time progress, and dark/light themes.
• Thermal & Power Monitoring: Active tracking of battery/CPU temperatures, power usage (watts), and battery drain rate during execution.

📋 Benchmark Tests (46 Total)

💻 CPU Tests (10 tests)

All tests run in both Single-Core and Multi-Core configurations:

• Prime Generation: Evaluates integer math using prime verification (Pollard's Rho).
• Fibonacci Iterative: Tests basic loops and iterative calculations.
• Matrix Multiplication: Tests floating-point and memory alignment throughput.
• Hash Computing (SHA-256/MD5): Measures cryptographic hash generation performance.
• String Sorting: Evaluates array manipulation and sorting algorithms (Quicksort).
• Ray Tracing: Measures floating-point and spatial computing speeds using Perlin Noise.
• Compression (LZMA): Tests data compression/decompression algorithms and memory throughput.
• Monte Carlo: Simulates mathematical approximations using randomized algorithms (Mandelbrot Set).
• JSON Parsing: Tests text processing, string manipulation, and object tree instantiation.
• N-Queens: Tests recursion, backtracking, and CPU branch prediction.

🧠 AI/ML Tests (5 tests)

• LLM Inference: On-device Large Language Model inference (llama.cpp with TinyLlama-1.1B/Gemma 3).
• Image Classification: Object recognition in images using ONNX Runtime (MobileNetV3).
• Object Detection: Identifies and bounds objects in real-time (EfficientDet Lite0).
• Text Embedding: Natural language vector generation (MiniLM-L6).
• Speech-to-Text: High-fidelity on-device audio transcription (Whisper-tiny).

🎮 GPU Tests (10 tests)

• Native GPU Tests:
  • Triangle Rendering: OpenGLES/Vulkan stress rendering pipelines.
  • Compute Shader Matrix: Heavy matrix multiplication run on GPU hardware.
  • Particle System: Simulation of 100K+ interactive physical particles.
  • Texture Sampling: High-resolution texture cache sampling and fillrate test.
  • Tessellation: Geometry subdivision and dynamic LOD rendering.
• External GPU Tests (Unity / Unreal):
  • Separate deep-linked high-fidelity rendering benchmark APKs (2 Unity scenes, 3 Unreal scenes) that communicate real-time FPS and frame times back to the main app.

💾 RAM Tests (5 tests)

• Sequential Read/Write: Measures block transfer speeds in system memory.
• Random Access Latency: Detects memory latency and page hopping speed.
• Memory Copy: Evaluates byte array copying throughput.
• Multi-threaded Bandwidth: Measures concurrent core access speeds.
• Cache Hierarchy: Tests L1/L2/L3 cache sizes and latency transitions.

📦 Storage Tests (6 tests)

• Sequential Read & Write: Benchmarks internal storage file reading and writing speeds.
• Random Read/Write: Tests storage IOPS with 4KB non-contiguous blocks.
• Small File Operations: Stress-tests storage index metadata speed with thousands of small files.
• Database Performance: Evaluates query/transaction speeds using local SQLite databases.
• Mixed Workload: Simulates simultaneous reading and writing activities on internal storage.

⚡ Productivity Tests (10 tests)

• UI & Dynamic Rendering: Measures RecyclerView scrolling, Canvas rendering speed, dynamic UI layout, and text/typography rendering operations.
• Media Workloads: Heavy image processing (batch resize, color filters) and video processing (H.264/H.265 encoding, transcoding).
• Document Workloads: High-page PDF document generation and rendering tests.
• Multi-tasking: Simulates heavy background workloads run concurrently with active UI animations.

📂 Benchmark Documentation

For detailed technical specifications, design documents, and implementation plans, check out the documentation under the /docs directory:

• Overview & Full Features
• Benchmark Index
  • CPU Benchmarks Details
  • AI/ML Benchmarks Details
  • GPU Native Benchmarks Details / GPU External Benchmarks Details
  • RAM Benchmarks Details
  • Storage Benchmarks Details
  • Productivity Benchmarks Details
• Database Schema & Architecture
• Benchmark Modes & Configurations

🔧 Building

git clone https://github.com/abhay-byte/finalbenchmark-platform.git
cd finalbenchmark-platform
./gradlew assembleDebug

📄 License

This project is licensed under the Apache 2.0 License - see the LICENSE file for details.

---

---