> ## Documentation Index
> Fetch the complete documentation index at: https://devdocs.xbox.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Minecraft case study

> Mojang's Minecraft Java FPS limiter for AFK and minimized states cuts power up to 75%, plus a console framerate cap on controller-disconnect dialogs.

**Studio:** Mojang<br />**Shipped:** August 2024 (Java snapshot)<br />**Contributors:** Nick Horvath (Engineering Systems Manager), Sue Loh (Performance Tech Lead), Timur Nazarov (Java), David Ekermann (Bedrock)

<Note>
  **Impact math:** A title averaging 160 W played 3 hours/day for a year consumes \~175 kWh per device. Across 100k devices, that equals GHG emissions from \~13.5M pounds of coal burned ([EPA calculator](https://www.epa.gov/energy/greenhouse-gas-equivalencies-calculator)). A modest 10% efficiency win at scale is substantial.
</Note>

Globally, PC gaming consumes \~230 TWh of electricity per year — equivalent to 59M EU households. In the US, gamers spend \~\$6B on electricity and emit \~12M tons of CO₂ annually. Mojang set out to reduce Minecraft's contribution to that footprint.

## Java: the "Reduce FPS when" setting

A new video setting introduces two modes:

* **Minimized** — caps framerate at 10 FPS when the window is minimized.
* **AFK (default)** — caps framerate at 30 FPS when no player input for over a minute, then further to 10 FPS after 10 minutes total. Also caps at 10 FPS when minimized.

Unlike a global FPS cap, this only engages when the player isn't actively playing.

### AFK mode measurements (NVIDIA GeForce RTX 3080, GPU-only)

| Version               | Gameplay | Idle \< 1 min | Idle 1–10 min | Idle > 10 min |
| --------------------- | -------- | ------------- | ------------- | ------------- |
| **1.21 (old)**        | 101 W    | 93 W          | 93 W          | 93 W          |
| **FPS Limiter (new)** | 100 W    | 82 W          | 32 W          | 25 W          |
| **OS idle baseline**  | 17 W     | 17 W          | 17 W          | 17 W          |

After 10 minutes of inactivity, Minecraft draws roughly what the OS itself needs — a **\~73% reduction** vs. the old idle baseline.

### Minimized measurements

| Version               | Gameplay | Minimized |
| --------------------- | -------- | --------- |
| **1.21 (old)**        | 101 W    | 134 W     |
| **FPS Limiter (new)** | 100 W    | 25 W      |

<Note>
  The old build actually drew **more** power when minimized (134 W vs. 101 W in gameplay) — the game continued to render offscreen at full framerate. The new build drops it to 25 W.
</Note>

## Console: controller disconnected framerate cap

Minecraft telemetry showed the **Controller Disconnected** dialog was on screen surprisingly often, and the game still rendered it at a full 60 FPS.

### Screen time by platform

| Device & area                               | Screen time | Median FPS |
| ------------------------------------------- | ----------- | ---------- |
| Windows                                     | 7.1%        | 60         |
| XBOX One: controller-disconnected screen    | 14%         | 60         |
| XBOX One: pause screen                      | 1%          | 60         |
| XBOX Series: controller-disconnected screen | 10.6%       | 60         |
| XBOX Series: pause screen                   | 0.5%        | 60         |

The team dropped the framerate on the controller-disconnected dialog to **10 FPS** after 60 seconds visible. Measured via PIX Power Monitor:

| Console           | Active gameplay | Stage 1 (60 FPS dialog) | Stage 2 (10 FPS dialog) |
| ----------------- | --------------- | ----------------------- | ----------------------- |
| **XBOX Series X** | 24% (102 W)     | 21% (89 W)              | 17% (81 W)              |
| **XBOX Series S** | 31% (62 W)      | 26% (52 W)              | 19% (48 W)              |

The experiment was validated and the change is being made permanent for the single-player scenario in an upcoming Bedrock release, with future work planned for multi-player scenarios and additional dialogs.

## Tools used

* **HWInfo** — PC GPU-only power measurement.
* **PIX Power Monitor** — XBOX console measurement.
* **Certification testing** — controlled repeatable capture.

## Takeaways for other studios

* **Instrument for wasted power first.** Telemetry showed the disconnected-controller dialog was on screen 10–14% of XBOX playtime.
* **Rendering something the player isn't looking at is pure waste.** Minimized windows, disconnected controllers, and long idle sessions are all candidates.
* **Stage your idle drops.** 30 FPS after 1 minute, 10 FPS after 10 minutes preserves responsiveness for players who return quickly.

## Related case studies

* [Fortnite & Unreal Engine](/build/game-principles/sustainability/case-studies/case-studies-fortnite)
* [Call of Duty](/build/game-principles/sustainability/case-studies/case-studies-cod)
* [The Elder Scrolls Online](/build/game-principles/sustainability/case-studies/case-studies-elder-scrolls-online)
* [Energy Efficiency Essentials](/build/game-principles/sustainability/energy-efficiency-essentials)


## Related topics

- [For Honor case study](/build/game-principles/sustainability/case-studies/case-studies-for-honor.md)
- [Star Wars Outlaws case study](/build/game-principles/sustainability/case-studies/case-studies-star-wars-outlaws.md)
- [Sustainability case studies overview](/build/game-principles/sustainability/case-studies/case-studies-overview.md)
- [Halo Infinite case study](/build/game-principles/sustainability/case-studies/case-studies-halo.md)
- [The Elder Scrolls Online case study](/build/game-principles/sustainability/case-studies/case-studies-elder-scrolls-online.md)