NVIDIA App Lets Users Split GPU Time

The shipped update lets owners of NVIDIA graphics cards choose how GPU time is divided between their active program and background tasks.

Updated on Nov. 4, 2026 in Computers

NVIDIA App Lets Users Split GPU Time

NVIDIA released version 11.0.9.585 of its NVIDIA App, adding a GPU resource prioritization tool that lets users decide how graphics processing time is shared between a foreground program and background tasks. The feature is a shipped software update, not an announced one.

Why it matters

The update moves GPU scheduling, previously largely invisible to end users, into a user-facing control panel. It signals a broader trend of consumer software exposing fine-grained hardware resource management once reserved for developers and system administrators.

The release is documented as a software specification: the default split of 80% foreground and 20% background GPU resources is set in the update, with four predefined profiles. Benchmark figures showing the measured effect on frame rates or background throughput are not reported in the source material.

The players

NVIDIA

A GPU maker whose hardware and CUDA software stack underpin most AI training and high-end graphics, and which distributes the NVIDIA App as its control software for GeForce cards.

The details

The tool sits in the Advanced tab of the NVIDIA App's System Settings and controls how GPU computing time is distributed between a program in focus and tasks running behind it. Users pick one of four presets: foreground application priority, standard mode, background task priority, or a custom configuration. The custom option opens a manual mode with a slider that sets the percentage of power allocated, so the system balances the GPU between competing workloads rather than leaving that decision to a fixed default.

Timeline

  1. Version 11.0.9.585 of the NVIDIA App was released on November 4, 2026.

The Tech Race

The release places GPU resource control in a consumer application while keeping performance counter access gated, with telemetry limited to system administrators by default. That split mirrors the developer-versus-user boundary NVIDIA maintains through its Nsight tooling.

Owners of NVIDIA GPUs running the NVIDIA App can open System Settings, go to the Advanced tab, and choose a priority profile or set a custom split. Availability beyond the NVIDIA App's current supported hardware is not specified in the source material.

The takeaway

Watch whether the 80/20 default changes measurably in independent benchmarks of foreground and background workloads. Readers who manage GPU-heavy systems can test the four profiles to see which split suits their workload.

Further reading

For broader context on how graphics hardware is managed in consumer machines, explore the Computers section.