NVIDIA announced Resource Management for RTX Spark on October 28, 2026. The feature lets users choose whether shared GPU compute favors a foreground app or a background-running agent. If an agent is working while you use another app, you can prioritize the app on screen or give the agent more resources to finish sooner.
What Resource Management controls
Resource Management sets priority for GPU compute, the processing work performed by the graphics processor. NVIDIA describes two choices: favor the foreground app for performance, or give more resources to the background agent. The choice addresses which task gets priority; it is separate from dividing memory between the GPU and the system.
For context on the platform’s DLSS 5 announcements, see NeoTeo’s earlier RTX Spark coverage.
Where to find the setting
NVIDIA lists the control under NVIDIA App > System > Advanced. It requires hardware-accelerated GPU scheduling to be enabled in Windows Settings.
Compute priority and memory allocation are separate controls
Resource Management governs which workload gets priority for GPU compute. A different NVIDIA App slider allocates RTX Spark’s unified physical memory between GPU-dedicated memory and CPU/system memory. The two controls appear under System > Advanced, but they change different things. NVIDIA’s support guidance says to restart after changing the memory allocation when prompted.
