How to Check GPU Usage on Mac

Open Activity Monitor (Cmd+Space, type "Activity Monitor," press Return) and navigate to the Window menu in the menu bar. Select GPU History to open a small graph showing GPU utilization over time. This graph provides a general sense of GPU activity but does not show utilization percentages, VRAM consumption, or per-process GPU usage.

For Apple Silicon Macs, the Terminal provides more granular data. Running sudo powermetrics --samplers gpu_power -i 1000 outputs GPU utilization percentage, active frequency, and power consumption at one-second intervals. This is the most detailed built-in method for checking GPU usage on Apple Silicon.

On Intel Macs with discrete GPUs, Activity Monitor's Energy tab shows which applications are using the high-performance GPU versus the integrated GPU. This helps identify apps that trigger the discrete GPU, which consumes significantly more power. The Activity Monitor guide covers the full interface in detail.

Apple menu > About This Mac displays the chip model for Apple Silicon Macs (such as M1, M2, M3, or M4). The GPU core count varies by chip variant. For example, the M3 base chip includes 10 GPU cores while the M3 Max includes up to 40 GPU cores. The Check Mac Specs guide walks through finding all hardware details.

For more detail, open System Report(click "More Info" in About This Mac, then "System Report") and navigate to Graphics/Displays. This section shows the GPU model name, VRAM allocation (for discrete GPUs on Intel Macs), Metal support version, and connected displays.

Apple Silicon uses a unified memory architecture. The GPU shares the same physical RAM as the CPU, so there is no separate VRAM pool. A Mac with 16 GB of unified memory allocates portions of that total to the GPU dynamically based on workload. This is fundamentally different from Intel Macs with discrete GPUs, which had dedicated VRAM (typically 4 GB or 8 GB) that was reserved exclusively for graphics.

The GPU History window in Activity Monitor updates in real time and provides a scrolling graph of GPU utilization. However, it must remain open as a visible window, which consumes screen real estate and competes with your active workspace. There is no built-in macOS option to pin GPU usage to the menu bar or Dock.

Third-party monitoring tools solve this by embedding GPU metrics directly in the macOS menu bar. This approach provides continuous visibility without requiring a dedicated window. Menu bar monitors typically display utilization percentages, temperature readings, and memory usage in a compact format that stays visible across all applications and workspaces.

For power users comfortable with the Terminal, powermetrics delivers detailed GPU telemetry including per-cluster utilization on Apple Silicon and power draw in milliwatts. The ioreg command can query GPU properties from the IOKit registry. Both tools output raw data that requires scripting to format into something human-readable. The Mac Performance Monitor guide covers additional monitoring approaches.

Video playback and encoding are among the most GPU-intensive tasks on a Mac. Hardware-accelerated video decoding (H.264, H.265/HEVC, ProRes) offloads work from the CPU to dedicated media engines on Apple Silicon, but encoding large video files still drives GPU utilization high for extended periods.

3D rendering and machine learning workloads push GPU cores to full utilization. Applications like Blender, Cinema 4D, and ML frameworks using Metal Performance Shaders or Core ML can sustain 100% GPU usage during compute-heavy operations.

Multiple external displays increase GPU compositor load. Each connected display requires the GPU to composite and render an additional framebuffer. Running two or three 4K displays simultaneously creates meaningful baseline GPU usage even without any GPU-intensive applications running.

Browser hardware acceleration offloads WebGL content, CSS animations, and video playback to the GPU. Websites with complex animations, WebGL visualizations, or multiple embedded videos can cause surprisingly high GPU utilization. Games, whether native macOS titles or browser-based, are also consistent sources of sustained GPU load.

MoniThor places GPU utilization front and center in the menu bar. The current utilization percentage updates continuously, letting you see GPU load at a glance while working in any application. A 60-sample sparkline graph sits alongside the percentage, showing recent GPU activity trends without requiring you to open a separate monitoring window.

When VRAM data is available, MoniThor displays used versus total memory dedicated to the GPU. On Apple Silicon Macs where the GPU shares unified memory with the CPU, MoniThor identifies the GPU by its proper Apple branding (such as "Apple M3 Pro") and reports memory allocation accordingly.

This persistent visibility eliminates the need to keep Activity Monitor open or run Terminal commands repeatedly. Whether you are encoding video, training a model, or simply curious about baseline GPU activity, the data is always one glance away.

Visit the Features page for the complete list of metrics, or return to the home page for an overview of MoniThor.

Close GPU-intensive applications. Games, 3D rendering software, and video editors consume the most GPU resources. Quitting these applications when they are not actively needed frees GPU capacity immediately.

Reduce external display count or resolution.Each external display adds compositor overhead. If you are experiencing high GPU usage at idle, try disconnecting a display or lowering its resolution in System Settings > Displays.

Disable browser hardware acceleration.In Chrome, navigate to Settings > System and toggle off "Use hardware acceleration when available." In Firefox, go to Settings > General > Performance and uncheck "Use recommended performance settings," then uncheck "Use hardware acceleration when available." This shifts rendering work from the GPU to the CPU, which may reduce GPU usage at the cost of slightly higher CPU load.

Use hardware-accelerated video codecs. Apple Silicon includes dedicated media engines for H.265/HEVC and ProRes. Encoding and decoding in these formats uses the media engine rather than the general GPU cores, resulting in lower GPU utilization and better energy efficiency compared to software-based codecs.

Check Activity Monitor's Energy tab. The Energy tab shows which applications are using the GPU. Sort by GPU usage to identify unexpected processes consuming GPU resources. The Mac CPU Usage guide covers how CPU and GPU demands interact when optimizing overall system performance.