Understanding GPU Bottlenecks

Lesson Length: 25–35 Minutes

Lesson Overview

In the previous lesson, we explored how the CPU calculates the simulation. In this lesson, we’re shifting our focus to the Graphics Processing Unit (GPU).

Many simmers assume that low FPS always means they need a better graphics card. While the GPU is extremely important, it’s only one part of the performance equation.

By the end of this lesson you’ll understand:

• What a GPU bottleneck is.
• How your graphics settings affect GPU workload.
• Why higher resolutions require significantly more processing power.
• What GPU utilisation means.
• How to identify when your GPU is the limiting factor.
• Which settings have the greatest impact on graphics performance.

What Is A GPU Bottleneck?

A GPU bottleneck occurs when your graphics card is working at its maximum capacity and cannot render frames any faster.

Unlike a CPU bottleneck, where the GPU is waiting for instructions, a GPU bottleneck means the graphics card is fully occupied rendering the image.

Typical signs include:

• High GPU utilisation (95–100%)
• Lower FPS when increasing graphics settings
• Stable CPU usage while GPU remains maxed out
• Better performance when lowering graphics settings or resolution

Example

Imagine an artist painting a picture.

The CPU hands the artist a new drawing to paint.

If the artist finishes quickly, they ask for the next one.

But if the drawing becomes extremely detailed, the artist takes longer to complete it.

The CPU is ready with the next frame, but the GPU is still busy drawing the previous one.

That is a GPU bottleneck.

Understanding Resolution

Resolution is one of the biggest factors affecting GPU performance.

Resolution determines how many pixels your graphics card must render every single frame.

Common resolutions include:

1920 × 1080 (1080p)

Approximately 2 million pixels per frame.

2560 × 1440 (1440p)

Approximately 3.7 million pixels.

Nearly twice as much work as 1080p.

3840 × 2160 (4K)

Approximately 8.3 million pixels.

Over four times as many pixels as 1080p.

Why Resolution Matters

Every pixel requires calculations.

Your GPU must determine:

• Colour
• Brightness
• Shadows
• Reflections
• Lighting
• Anti-aliasing
• Texture detail

Now imagine doing this 60 times every second.

Higher resolution equals dramatically more GPU work.

Example

At 1080p

Your RTX 4070 might achieve:

80 FPS

At 4K

The same graphics card might achieve:

40–50 FPS

The graphics card hasn’t become slower.

It’s simply doing much more work.

Texture Resolution

Textures are the detailed images applied to aircraft, airports and scenery.

Examples include:

Cockpit panels

Runways

Taxiways

Buildings

Aircraft paint

Ground markings

Roads

Trees

Cloud textures

Higher texture settings create sharper images.

However…

They require more VRAM.

If your graphics card doesn’t have enough VRAM…

You may experience:

Texture pop-in

Long loading times

Micro-stutters

Reduced FPS

Recommended Approach

If your graphics card has:

8GB VRAM

Use High textures carefully with demanding add-ons.

12GB+

High or Ultra textures are usually more comfortable.

Remember: available VRAM isn’t the only factor—your overall settings and scenery also matter.

Clouds

Microsoft Flight Simulator features volumetric clouds.

Unlike flat cloud textures, volumetric clouds are three-dimensional.

The GPU calculates:

Cloud depth

Lighting

Sunlight

Cloud shadows

Weather effects

Storms

Fog

Every frame.

Cloud quality is one of the most GPU-intensive settings in the simulator.

Reducing cloud quality often provides a noticeable improvement in GPU-limited scenarios.

Ambient Occlusion

Ambient Occlusion creates realistic contact shadows where objects meet.

Examples:

Cockpit switches

Instrument panels

Seats

Landing gear

Buildings

Ground objects

Without Ambient Occlusion…

Everything appears flatter.

With Ambient Occlusion…

Objects gain depth and realism.

Because these shadows are recalculated constantly, higher settings increase GPU workload.

Reflections

Reflections add realism to surfaces such as:

Cockpit windows

Aircraft fuselage

Water

Buildings

Wet runways

Glass terminals

The GPU must calculate:

What is being reflected

Where it’s reflected

How strong the reflection should appear

The more realistic the reflection…

The greater the GPU workload.

Anti-Aliasing

Without Anti-Aliasing…

Straight lines appear jagged.

Examples include:

Wing edges

Buildings

Taxiway lines

Power cables

Runways

Anti-Aliasing smooths these edges.

Microsoft Flight Simulator currently offers several methods.

TAA

Produces excellent image quality.

Higher GPU workload.

DLAA

Excellent clarity.

Higher GPU workload.

DLSS

Uses AI to render at a lower internal resolution before reconstructing the image.

Benefits:

Higher FPS

Reduced GPU workload

Very effective on supported NVIDIA RTX cards.

Trade-offs:

Some users may notice slight softness or ghosting depending on the version and settings.

FSR

AMD’s equivalent technology.

Uses upscaling to improve performance on supported hardware.

DLSS Explained

DLSS stands for:

Deep Learning Super Sampling.

Instead of rendering every pixel at full resolution…

The GPU renders fewer pixels.

Artificial Intelligence then reconstructs the image.

This significantly reduces GPU workload while maintaining impressive visual quality.

This is why many RTX users see substantial FPS improvements with DLSS enabled.

Frame Generation

Frame Generation is available on supported NVIDIA RTX 40-series graphics cards.

Instead of rendering every frame…

Artificial Intelligence generates additional frames between the real ones.

Example:

Without Frame Generation:

60 FPS

With Frame Generation:

100+ FPS

This can greatly improve perceived smoothness in GPU-limited situations.

However, it doesn’t reduce CPU workload. If your simulator is heavily CPU-limited, Frame Generation can improve the visual fluidity but won’t remove the underlying CPU limitation.

Understanding GPU Utilisation

GPU utilisation measures how much of your graphics card’s processing capacity is currently being used.

Example:

30%

The graphics card has plenty of unused capacity.

60%

Moderate workload.

95–100%

The graphics card is working near its maximum capacity.

This is expected in a GPU-limited scenario.

When High GPU Usage Is Good

Many users worry when they see 99% GPU usage.

In reality…

If your CPU is keeping up…

This usually means your graphics card is being fully utilised.

That’s often exactly what you want.

When Low GPU Usage Is A Problem

Example:

GPU Usage

55%

FPS

32

Why isn’t the GPU working harder?

Because the CPU hasn’t finished preparing the next frame.

The GPU is waiting.

This usually indicates a CPU bottleneck.

GPU-Limited Scenarios

Let’s compare three common scenarios.

Scenario 1 – Flying Over Rural Countryside

1080p

Medium settings

GPU Usage

55%

CPU does most of the work.

Excellent FPS.

Scenario 2 – Flying Into Heathrow At 4K

Ultra settings

Heavy clouds

RTX 4070

GPU Usage

99%

The graphics card is fully occupied.

Lowering cloud quality or reflections would likely improve FPS.

Scenario 3 – Ultra-Wide Monitor

3440 × 1440

Ultra textures

Maximum reflections

GPU Usage

98–100%

Large numbers of pixels combined with demanding graphics settings create a GPU bottleneck.

Student Exercise

Compare Different Resolutions

If you have access to multiple resolutions:

Step 1

Enable the Developer Mode FPS Counter.

Step 2

Load into the same airport using the same aircraft and weather.

Step 3

Record your results at 1920 × 1080.

Write down:

• Average FPS
• GPU utilisation
• Main Thread time
• GPU frame time

Step 4

Repeat the test at 2560 × 1440 or 3840 × 2160 (if your monitor supports it).

Step 5

Compare the results.

Ask yourself:

• Did GPU utilisation increase?
• Did FPS decrease?
• Did the image quality improve enough to justify the performance cost?
• Was the GPU or the Main Thread the limiting factor?

If your monitor only supports one resolution, you can still complete this exercise by changing a GPU-heavy setting (such as Cloud Quality or Reflections) and observing how GPU utilisation and FPS respond.

Lesson Summary

Congratulations.

You now understand how the GPU affects Microsoft Flight Simulator performance.

Remember:

• Resolution is one of the largest GPU workloads.
• Texture quality mainly affects VRAM usage.
• Volumetric clouds are among the most demanding graphics settings.
• Ambient Occlusion and Reflections improve realism but increase GPU load.
• Anti-Aliasing smooths image quality at a performance cost.
• DLSS can improve performance by reducing rendering workload.
• Frame Generation increases perceived smoothness on supported hardware but doesn’t remove CPU bottlenecks.
• GPU utilisation helps you determine whether your graphics card is the limiting factor.