Benchmarks 5 min read

AnTuTu vs Standard APK: Can Benchmark Spoofing Prevent Thermal Throttling in Android PC Emulation?

By DroidDeck Engineering Lab •

When executing PC games through DroidDeck, your device's Qualcomm processor performs two demanding tasks at the exact same time: translating x86/x64 instructions to ARM64 via FEX, while rendering DirectX 11/12 graphics commands on the Adreno GPU via Mesa Turnip.

Under this combined workload, a smartphone's system-on-chip generates between 8 and 14 watts of heat.

On stock Android firmware, the operating system's internal thermal monitoring daemon—Qualcomm's thermal-engine—monitors thermistor sensors across the battery, display, and SoC. When standard applications cross manufacturer temperature thresholds, clock speeds are cut by as much as 40% to 55%.

To evaluate whether DroidDeck's AnTuTu Benchmark Spoof (com.antutu.benchmark.full) can prevent this thermal degradation, our engineering lab put both builds through an intensive 30-minute stress test. Here are our findings.

---

1. How Qualcomm's Thermal Engine Treats Benchmark Packages

To understand why the AnTuTu variant performs differently, we must examine Android's Hardware Abstraction Layer (HAL) and vendor thermal architecture:

bash
Android Thermal Protection Layers:
┌────────────────────────────────────────────────────────┐
│  Qualcomm thermal-engine / mithermald / secthermal   │
│                                                        │
│  [Standard Package: com.droiddeck.launcher]            │
│  ├─ Skin Temp Threshold: 39°C  --> Throttling Step 1   │
│  ├─ Skin Temp Threshold: 42°C  --> Throttling Step 2   │
│  └─ Core Clocks: Dropped to 350-490 MHz (Stuttering)   │
│                                                        │
│  [Benchmark Package: com.antutu.benchmark.full]        │
│  ├─ Skin Temp Threshold: 46°C  --> High Threshold      │
│  ├─ Thermal Trip Point:  85°C core temp before downclock│
│  └─ Core Clocks: Sustained 719-903 MHz (Smooth 60 FPS) │
└────────────────────────────────────────────────────────┘

When smartphone OEMs submit phones for review, benchmark apps like AnTuTu and Geekbench are expected to push silicon to its maximum limits. Manufacturers therefore program their thermal rules to raise trip points specifically for com.antutu.benchmark.full:

  • Thermal downclocking triggers are delayed until battery temperature reaches ~45°C instead of ~39°C.

  • The GPU governor (msm-adreno-tz) is forced into its top performance frequency tiers (pstate 0 and pstate 1).

  • Core affinity locking keeps the ARM Cortex-X3 / Cortex-X4 prime cores engaged longer.
  • For a full breakdown of all four available builds, consult our DroidDeck APK Variants Comparison Guide.

    ---

    2. Benchmark Test Methodology & Hardware Setup

  • Test Device: Xiaomi 14 (Qualcomm Snapdragon 8 Gen 3, Adreno 750, 16 GB LPDDR5X, HyperOS 1.0).
  • Ambient Room Temperature: 22°C (71.6°F) controlled environment.
  • Display Output: 1080p native panel, 120Hz refresh mode.
  • Thermal Measurement: Internal SoC thermistors via cat /sys/class/thermal/thermal_zone*/temp + FLIR infrared camera for external chassis skin temperatures.
  • Telemetry & Overlay: In-engine MangoHud v0.7.2 compiled for ARM64 Proton.
  • Game Tested: *Cyberpunk 2077* (Patch 2.13, 720p Low settings, Turnip v24.3-devel, internal gamescope integer scaling).
  • ---

    3. Sustained Performance Over 30 Minutes: Standard vs AnTuTu

    Here is the minute-by-minute performance log recorded during our test session:

    Benchmark TimeStandard APK (com.droiddeck.launcher)AnTuTu Spoof (com.antutu.benchmark.full)Delta (Improvement)
    0 – 5 Minutes58.4 FPS (Adreno 750 @ 903 MHz)59.2 FPS (Adreno 750 @ 903 MHz)+1.3% (Margin of error)
    10 Minutes51.2 FPS (Adreno 750 @ 770 MHz)58.6 FPS (Adreno 750 @ 903 MHz)+14.4%
    15 Minutes42.1 FPS (Adreno 750 @ 615 MHz)57.1 FPS (Adreno 750 @ 770 MHz)+35.6%
    20 Minutes36.8 FPS (Adreno 750 @ 495 MHz)54.3 FPS (Adreno 750 @ 770 MHz)+47.5%
    25 Minutes33.5 FPS (Adreno 750 @ 410 MHz)52.8 FPS (Adreno 750 @ 680 MHz)+57.6%
    30 Minutes31.9 FPS (Adreno 750 @ 380 MHz)51.4 FPS (Adreno 750 @ 680 MHz)+61.1%
    bash
    Framerate Stability Across 30 Minutes:
    60 FPS ┌──────────────────────────────────────────────┐
           │AnTuTu:  ████████████████████████████████░░░░░│ (Avg: 55.6 FPS)
    30 FPS │Standard:██████████░░░░░░░░░░░░░░░░░░░░░░░░░░░│ (Avg: 42.3 FPS)
     0 FPS └──────────────────────────────────────────────┘
           0 min         10 min         20 min         30 min
    

    Key Observations:

  • The Standard APK throttled at Minute 8: Once the battery thermistor reached 39.5°C, HyperOS reduced GPU clock speeds from 903 MHz down to 615 MHz, causing noticeable stuttering during gunfights in Night City.
  • The AnTuTu APK sustained peak clocks for 18 minutes: The GPU did not downclock below 770 MHz until Minute 19, maintaining a fluid 50+ FPS experience even in heavy combat.
  • 1% Low Framerates Improved by 82%: The AnTuTu build practically eliminated micro-freezes caused by rapid GPU frequency stepping.
  • ---

    4. Thermals and Battery Consumption Comparison

    Allowing the processor to sustain higher clock speeds produces more heat and consumes more battery power:

    Telemetry MetricStandard EditionAnTuTu Spoof Edition
    Peak SoC Temperature68°C81°C
    Peak Rear Glass Temperature41.2°C (Warm)46.8°C (Hot to touch)
    Battery Drain (30 mins)14% (~3.5 hrs total runtime)22% (~2.2 hrs total runtime)
    Average Power Draw7.8 Watts12.1 Watts
    ---

    5. Engineering Recommendations & Thermal Safety

    While the AnTuTu spoof variant provides massive framerate improvements on aggressively throttled devices, you must take reasonable thermal precautions:

  • Do Not Play While Charging on Fast-Charge Bricks: Fast charging generates 5–10W of battery heat on its own. Combining fast charging with AnTuTu benchmark mode can push battery temperatures past 48°C, which triggers emergency OS shutdowns.
  • Use an Active Magnetic Semiconductor Cooler: If you frequently play heavy 3D titles (*Cyberpunk 2077*, *The Witcher 3*, *GTA V*) for longer than 30 minutes, an active Peltier cooler (e.g. Black Shark MagCooler or Flydigi B6X) keeps the rear glass under 30°C, allowing the AnTuTu build to maintain maximum 903 MHz GPU clocks indefinitely.
  • Dedicated Handheld Owners Don't Need This: If you own an Ayn Odin 2 or Ayaneo Pocket S, your device already has an internal cooling fan and unlocked thermal governors. The Standard APK will run at full speed without benchmark spoofing. Read our Handhelds vs Smartphones Guide for details.
  • ---

    Conclusion

    Our benchmarks demonstrate that package name spoofing is not a gimmick—it produces measurable, dramatic performance gains on commercial Android phones. By switching to the AnTuTu variant, you bypass arbitrary battery-saver governors and unlock the full raw computing power of your Snapdragon hardware.

    Grab the AnTuTu APK variant directly from the DroidDeck Download Portal, or explore the guides below to fine-tune your mobile PC gaming rig.

    ---

  • DroidDeck APK Variants Explained: Standard vs PUBG vs AnTuTu vs Ludashi
  • How to Bypass Samsung GOS & Xiaomi Game Turbo Throttling in DroidDeck
  • How to Optimize Snapdragon 8 Gen 2 & Gen 3 for Sustained 60 FPS in DroidDeck
  • Android Gaming Handhelds vs Smartphones: Why Handhelds Don't Need Spoofing
  • Can You Run GTA V on Android? DroidDeck Performance & Settings
  • Get Started on Your Android Device

    Experience SteamOS on Your Handheld Today

    Ready to run Valve's official Steam client in Big Picture mode with ARM64 Proton? Download the verified DroidDeck APK and verify your device compatibility in minutes.

    Related Articles in this Cluster