2 Sources
[1]
Free open-source GPU task manager reveals exactly what's eating your VRAM -- Microsoft VP's new GPU Taskman tool reveals every process using your GPU, how much memory each uses, and breaks down AI and LLM workloads
Open source tool is available for Windows, Linux, and macOS. Microsoft VP Scott Hanselman has revealed a new open-source GitHub project called gputaskman. The name neatly encapsulates the headline role of this tool, a GPU-focused task manager, but its scope might be broader than you expect. This
[2]
Microsoft developer builds a Task Manager app for your GPU
A Windows engineer develops 'gputaskman', a Task Manager-type app designed to list what processes are running on the GPU rather than the CPU. Task Manager is the go-to application in Windows for monitoring hardware utilization and hunting down bloated processes that are slowing down your PC.
Share
Copy Link
Microsoft VP Scott Hanselman has launched gputaskman, an open-source GPU task manager for Windows, Linux, and macOS. The tool reveals per-process GPU usage, VRAM consumption, and provides detailed breakdowns of AI and LLM workloads including model weights and architecture.
Microsoft VP Scott Hanselman has released gputaskman, an open-source GPU task manager that addresses a critical gap in system monitoring tools
1
2
. Available for Windows, Linux, and macOS, this Terminal-based application delivers granular insights into GPU activity that Windows Task Manager cannot provide. Hanselman explained his motivation succinctly: "There's a lot going on in my GPU. I'm making a GPU Taskman to get insights to where my GPU memory is going."1

Source: Tom's Hardware
The tool displays per-process GPU usage, VRAM consumption, active runtimes, GPU utilization, clock speeds, temperatures, and page file usage in a three-section interface
1
2
. The top window features hardware monitors including VRAM usage charts. The middle section lists processes with their runtime details and memory footprint. The bottom window reveals detailed app information when users select a process. Screenshots from Hanselman's RTX 4080S system showed the desktop windows manager consuming the most VRAM, alongside AI and LLM workloads and standard Windows applications1
. Live-streaming charts track GPU load and temperature in real-time, providing power users with the visibility they need to monitor GPU processes effectively.What distinguishes this open-source GPU task manager is its model inspector capability, specifically designed for AI and LLM workloads
1
2
. When llama-server.exe was selected in demonstration screenshots, the tool identified the Qwen3-0.6B model and broke down critical details including model weights, KV cache, parameter count, architecture, quantization, and layer count1
2
. This level of detail proves invaluable as the computing industry increasingly processes data on GPUs rather than CPUs, particularly for users running multiple AI models simultaneously.Related Stories
Hanselman released gputaskman version 0.3.0 under the permissive MIT license, making the source code and portable Windows downloads freely available on GitHub
1
. The developer describes it as "experimental software" but hints at broader implications. In social media comments, Hanselman suggested that expanded GPU monitoring capabilities could be integrated into Windows as standard1
. If gputaskman gains traction, it could incentivize Microsoft to build more comprehensive monitoring systems in Windows Task Manager2
. For now, users seeking to monitor GPU processes with unprecedented detail have access to a tool that fills a significant void in system monitoring capabilities.Summarized by
Navi
05 Nov 2026•Technology
24 Aug 2026•Technology
09 Dec 2025•Technology

1
Technology

2
Technology

3
Policy and Regulation
