How to Set Your Preferred GPU on a Laptop

Why a preferred GPU matters for programming

Many laptops come with two graphics processors. The CPU includes an integrated GPU, and a discrete GPU such as an NVIDIA GeForce RTX chip handles heavier graphics work. Windows can assign each app to either one, and that assignment is called the preferred GPU.

For programming, the correct choice depends on the task. An editor, a browser and a local database run well on integrated graphics. Machine learning, GPU-accelerated rendering and game development benefit from the discrete GPU. The goal is to let ordinary developer tools use less power and to reserve the discrete GPU for work that actually needs it.

If you are deciding whether your next laptop needs a discrete GPU at all, start with integrated vs dedicated graphics and then check the best laptops for programming.

Find the GPUs in your system

Open Task Manager in Windows and select the Performance tab. You will see one or more GPU entries, usually listed by vendor name. An integrated GPU is often labeled with the CPU brand, while a discrete GPU may appear as an NVIDIA GeForce device.

The Intel processor naming guide explains that select Intel Core Ultra processors can include an Intel Arc GPU for graphics and AI acceleration. That GPU is integrated into the processor package. The discrete GPU on a current laptop may be a GeForce RTX 50 Series chip, as described on NVIDIA's laptop page.

If you see only one GPU, that is the GPU Windows will use for everything. On Apple silicon, the GPU is part of the same chip and shares the same memory; for example, the MacBook Air tech specs list an Apple M5 chip with an 8-core or 10-core GPU.

Set a per-app GPU in Windows

Windows lets you set a preferred GPU for each app. This is more precise than a system-wide switch and lets you keep your editor on integrated graphics while sending heavier tools to the discrete GPU.

Open Settings, go to System, then Display, then Graphics. Choose an app from the list, select Options, and pick either the power saving option, which usually maps to the integrated GPU, or the high performance option, which maps to the discrete GPU. If an app is not listed, browse to its executable file and add it.

Use this setting for your editor if you want it to stay on integrated graphics. Use it for machine learning tools, renderers and game engines when you want those apps to use the discrete GPU.

Use the GPU vendor's control software

GPU vendors also provide control software for per-app graphics settings. NVIDIA's RTX laptop page describes the NVIDIA app with Game Ready and Studio Drivers, which is the natural place to set driver-level preferences for GeForce RTX laptops.

On systems where the integrated GPU is Intel Arc graphics, Intel's graphics control software offers similar per-app options. The Intel processor page notes that Intel Arc graphics is available on select Intel Core Ultra processors.

Vendor tools are useful when you need more control, such as forcing a specific GPU for a command-line tool or a game engine. Some laptops also include a MUX switch to select which GPU drives the internal display; see laptop mux switch for the difference.

What about Apple silicon

On Apple silicon MacBooks, there is no per-app preferred GPU setting in the same way. The GPU is part of the same chip as the CPU and uses the same unified memory pool. The MacBook Air tech specs show the M5 chip with an 8-core or 10-core GPU, and the system manages graphics work automatically.

For iOS and macOS development, the operating system requirement is the first filter. Xcode requires a supported macOS version, so you choose Apple hardware before worrying about GPU assignment. See best laptops for iOS and macOS development for guidance.

When to force the discrete GPU

For ordinary coding, the integrated GPU is usually enough. Visual Studio Code lists 1.6 GHz or faster processor and 1 GB of RAM as hardware recommendations, so the editor does not require a discrete GPU.

Force the discrete GPU for work that is explicitly GPU-bound: local machine learning, training or inference with local models, GPU-accelerated rendering, and game development previews. For those tasks, graphics memory often matters more than the GPU preference setting; see how much VRAM you need and GPU for machine learning.

If you work mainly in web development, you can leave the browser and editor on integrated graphics and still have a capable development machine. Start with web development laptops to match the rest of the specs to your workflow.

What to pick for your work

Questions

How do I know if my laptop has two GPUs?

Open the Performance tab in Task Manager and look for multiple GPU entries. An integrated GPU appears with the CPU vendor name, and a discrete GPU may appear as GeForce. If only one GPU is listed, that is what all apps use.

Does setting a preferred GPU speed up code compilation?

No. Compilation mostly uses the CPU and memory. The GPU preference affects graphics work and GPU-accelerated tasks, so keeping your IDE on integrated graphics is usually the right default.

Can I force every app to use the discrete GPU?

Yes, but you generally should not. Developer tools that do not need GPU acceleration run well on the integrated GPU. Reserve the discrete GPU for machine learning, rendering, game previews and similar GPU-bound work.

Does Apple silicon have a preferred GPU setting?

No. On Apple silicon the GPU is on the same chip and uses unified memory, so the system manages GPU work automatically. The MacBook Air tech specs show the GPU as part of the M5 chip.

Is the integrated GPU enough for Visual Studio Code?

Yes. Visual Studio Code recommends a 1.6 GHz or faster processor and 1 GB of RAM, so a discrete GPU is not required for the editor itself.

Should Docker Desktop use the discrete GPU?

Containers generally depend on CPU, memory and storage rather than the GPU. Keep container tooling on integrated graphics and spend your upgrade budget on RAM. See best laptops for Docker and virtual machines for the rest of the buying picture.