Get Started With Nsight Systems

Download NVIDIA Nsight Systems

Nsight Systems 2025.2.1 is Available Now

Review the supported platforms for NVIDIA Nsight™ Systems to choose the correct version for your host and profiling target.

If profiling from the CLI, pick your platform based on where the CLI will be run. If using the GUI (Full Version) to view reports, do profiling, or do remote profiling, pick your platform based on the host PC architecture where the GUI will be run.

Also review the system requirements before downloading.

Desktop, workstation, and server platforms:

Kubernetes integration:

JupyterLab integration:

Embedded and automotive platforms:

View Nsight Systems documentation.

---

Supported Platforms

Nsight Systems is distributed through multiple packages. Pick a “Profiling Target” column and learn what hosts may be used to profile (local or remote) as well as view reports.

		Profiling Target
		Linux Workstations & Servers	Windows Workstations & Gaming PCs	NVIDIA DPUs & SuperNICs	Jetson & IGX	DRIVE
From Host
	Windows	Remote GUI* Report Viewer**	Local CLI & GUI Remote GUI* Report Viewer**	Remote CLI Remote Report Viewer	Remote Report Viewer	Remote Report Viewer
	Mac	Remote GUI* Report Viewer**	Remote Report Viewer**	Remote Report Viewer	Remote Report Viewer	Remote Report Viewer
	Linux	Local CLI & GUI Remote GUI* Report Viewer**	Remote GUI Report Viewer**	Remote CLI Remote Report Viewer	Remote GUI Report Viewer***	Remote GUI Report Viewer***
	DPU / SuperNIC	N/A	N/A	Local CLI	N/A	N/A
	Jetson	N/A	N/A	N/A	Local CLI & GUI Report Viewer***	N/A
	DRIVE	N/A	N/A	N/A	N/A	Local CLI

* For x86-64 targets only or opening report collected from a CLI

** Only for reports collected from Windows or Linux PCs & servers of equal or lesser versions

*** Only for reports collected from Jetson or DRIVEOS of equal or lesser versions

---

System Requirements

Nsight Systems is compatible on Windows workstations and PCs, Linux workstations and servers, as well as Jetson and NVIDIA DRIVE Autonomous Machines. Learn about the system requirements and support for your development platform below.

	Windows Workstations and Gaming PCs	Linux Workstations and Servers	Linux Arm Servers	Jetson and Drive Autonomous Machines
Operating Systems	Windows 10 or newer	Ubuntu 24.04, 22.04, and 20.04* WSL-Ubuntu 2.0 CentOS 7+* RHEL 7, 8, 9 SLES 15 Debian 10, 11, 12 Fedora 37 KylinOS 10 OpenSUSE 15 Rocky 8, 9	Ubuntu, 20.04*, and 22.04 Rhel 8, 9 SLES 15	Jetson Linux DRIVE OS
Target Hardware	GPU: Pascal or newer CPU: x86-64 processors	GPU: Pascal or newer CPU: x86-64 processors**	GPU: Pascal or newer Arm-SBSA servers	NVIDIA IGX, Jetson AGX Orin, Jetson AGX Xavier, Jetson TX2, Jetson TX1, DRIVE AGX Orin, DRIVE AGX Pegasus, DRIVE AGX Xavier, DRIVE PX Parker AutoChauffeur, DRIVE PX Parker AutoCruise
Target Software	64-bit applications only CUDA 10.0+ for CUDA trace Driver 418 or newer***	64-bit applications only CUDA 10.0+ for CUDA trace Driver 418 or newer***	64-bit applications only CUDA 10.0+ for CUDA trace Driver 418 or newer***
Local Profiling	CLI and GUI	CLI and GUI	CLI and GUI	CLI (all platforms), GUI (Jetson Linux only)
Remote Profiling From Platforms	Windows 10+ macOS 11+ Ubuntu 20.04+	Windows 10+ macOS 11+ Ubuntu 20.04+	N/A	Ubuntu 22.04

* For older OS versions, please use Nsight Systems 2020.3
** Intel Haswell architecture or newer is required for LBR sampling backtrace
*** Driver 535 and newer improves GPU profiling stability. Please use the latest driver for the best results. Download here.

---

Release Notes

2025.2.1

• Highlights:
  • Dask API trace
  • PyTorch enhancements
  • Python 3.13 support
  • CUDA trace enhancements
    • (Beta) Hardware based low-overhead CUDA trace for NVIDIA Blackwell (--trace=cuda-hw)
    • GPU Direct Storage trace (--trace=gdc)
    • CUDA device side event trace (--cuda-event-trace)
    • Graph trace improvements
    • Kernel CGA dimensions & policy
    • Stream priority in tooltips
    • NVIDIA Confidential Compute support improvements
  • Windows graphics trace enhancements
    • GPU Frame Duration for DLSS Frame Generation
    • GPU resource trace tracks pre-start allocation names
    • Graphics Hotspot Analysis recipe
  • Linux system trace enhancements
    • Syscall trace enhancements (requires CAP_BPF and CAP_PERFMON)
      • Support system-wide mode (--syscall=pid-namespace)
      • Collect backtraces
    • OS Runtime Trace (OSRT) VFS POSIX functions trace (--osrt-file-access=true)
  • NVIDIA Grace support enhancements
    • Topdown analysis recipe for PMU events based on NVTX range annotations
    • Updates to available counters & metrics
  • NVIDIA Tools Extensions(NVTX) API & support enhancements
    • Various bug fixes
    • Payloads Extensions
    • Counters Extensions
    • Deferred Events Extensions
    • Updates bundled within NVIDIA Nsight Systems and github
  • NVIDIA Nsight Systems Plugins
    • Callback for last-chance to submit NVTX deferred events on stop
    • Windows support
  • GUI improvements
    • macOS GUI now available for arm64
    • Go to range — a timeline toolbar to quickly jump to the longest, shortest, and median ranges
  • NVIDIA Nsight Streamer is now available on NGC for viewing reports on remote headless servers
    • Docker
    • Kubernetes
  • NVIDIA Nsight Operator releasing soon on NGC for Kubernetes
    • Learn more here and apply for early access features

2025.1.1

• Highlights:
  • CUDA 12.8 support
  • Keep last N seconds CLI option - to retain the most relevant data when trying to record an unpredictable event, use --keep=N option with the ‘nsys stop’ command
  • Recipes
    • Summary of GPU metrics samples per range (NVTX or CUDA kernels)
  • Pytorch
    • Command-line option for enabling pytorch autograd layer NVTX ranges
    • Command-line option to trace prominent pytorch python functions
  • Windows
    • graphics resource tracker now includes resource priority changes
    • Reduced memory overhead when generating reports that contain ETW data
  • Python 3.12 support for scripts & recipes
  • Linux self-unpacking .run installer is now available for Arm64 SBSA systems
  • Preview: Multi-pass script to run all CPU counters involved in Arm Top-Down analysis

2024.7.1

• Highlights:
  • Storage profiling functionality expanded
    • Lustre and NFS volumes performance metrics on the client
    • Local disks and NVMe-oF volumes performance metrics on the client
    • Recipe for storage utilization per volume:
      • Cumulative throughput line graphs
      • Throughput heatmaps
    • Statistical analysis of MPI communication parameters
    • Installer package variant added for NVIDIA Grace - .run
    • Option to keep only last N seconds when generating a report file
    • Option to export a subsection of the timeline into a new report file
    • Enable "Resolve Symbols when creating report" by default on Windows and resolve only the symbols of the most relevant process

2024.6.1

• Highlights:
  • DirectX12 memory resource trace improvements:
    • New chart displays the distribution of DirectX12 resource types allocated by the target process in VRAM, across the trace session.
  • Storage profiling (beta)
    • Lustre & NFS - Performance metrics and operation counters for mounted volumes. (beta)
    • NVMe-oF (NVMe over fiber) - Performance metrics from mounted volumes. (beta)
    • Local storage - Performance metrics from local devices (beta)
    • Example use-cases
      • Verify GPU compute & storage transfer concurrency
      • Identify GPU compute idle due to a storage transfer dependencies

2024.5.1

• Highlights:
  • GPU metrics enhancements
    • Compute triage for NVIDIA H100 (preview)
    • Sync and async copy engine activity
  • Python call-stack sample statistics tables
  • OS System-Call trace (beta)
  • GPU power metrics sampler (preview)
  • NFS metrics sampler (beta)
  • Net Interface metrics sampler and plugin example code (beta)
  • CUDA trace support for devices with attribute cudaDevAttrD3D12CigSupported
  • New analysis recipes
    • NVIDIA NIC traffic statistics
    • MPI communication parameters statistics
  • Windows memory resources trace improvements
    • System-wide committed VRAM timeline chart
    • Additional memory transfer states: pending and in progress
    • Vulkan device memory object names
    • Vulkan bound resources names
  • Quality of life improvements
    • Windows symbols resolver now faster & limited to relevant processes
    • Windows ETL import now includes page fault events
    • Graphics frame health tooltips add thread IDs of API calls
• Additional Details:
• GPU metrics enhancements
  A new metrics-set is introduced for NVIDIA H100, focused on compute triage. This new set contains far more metrics. As such it must be collected at lower speeds or for a shorter duration otherwise the buffers will overflow. Copy engines are also exposed in the general metrics-set to better understand GPU activity for some architectures such as NVIDIA Ada Architecture. Synchronous copy engines are used in the graphics command sequences. Async copy engines are used in both compute and graphics to copy resources (typically) in the background.
• Python call-stack sample statistics tables
  Initial Python call-stack sampling support was only presented on the timeline as markers. Historically call-stack samples of C, C++, ELF binaries additionally were presented in function statistics tables to explore the frequency of function calls and stack paths. These same views are now available for Python as well.
• OS System-Call trace (beta))
  An alternative to OS RunTime API trace (OSRT) which is system-wide. OS runtime APIs are typically implemented with system calls (syscall instructions). System call trace collects syscall events from kernel space. Sudo or CAP_SYS_ADMIN is typically required.
• GPU power metrics sampler (preview)
  Low frequency sampling of GPU power and temperatures, for users interested in understanding when clocks are adjusted by temperature or optimizing for power.
• NFS metrics sampler (beta)
  See how network file system activity may relate to idle GPU time. Accessing data over the network just-in-time can cause delays vs prefetching it.
• Net Interface metrics sampler and plugin example code (beta)
  Adding upon existing networking support for NVIDIA NIC and AWS EFAs, this samples the throughputs on linux net interfaces. These can be physical devices or virtual adapters for mechanisms such as containers and virtual machines. IO patterns are sometimes on the critical path of computation instead of being asynchronous, or not started early enough to avoid blocking the critical path.

---

Feature Table

Feature	Linux Workstations and Servers	Windows Workstations and Gaming PCs	Jetson Autonomous Machines	DRIVE Autonomous Vehicles
View system-wide application behavior across CPUs and GPUs
CPU cores utilization, process, & thread activities	yes	yes	yes	yes
CPU thread periodic sampling backtraces	yes*	yes	yes	yes
CPU thread blocked state backtraces	yes**	yes	yes	yes
CPU performance metrics	yes	no	yes	yes
GPU workload trace	yes	yes	yes	yes
GPU context switch trace	yes	yes	yes	yes
SOC hypervisor trace	-	-	-	yes
SOC memory bandwidth sampling	-	-	yes	yes
SOC Accelerators trace	-	-	Xavier+	Xavier+
OS Event Trace	ftrace	ETW	ftrace	ftrace, QNX kernel events
Investigate CPU-GPU interactions and bubbles
User annotations API trace NVIDIA Tools Extension API (NVTX)	yes	yes	yes	yes
CUDA API	yes	yes	yes	yes
CUDA libraries trace (cuBLAS, cuDNN & TensorRT)	yes	no	yes	yes
OpenGL API trace	yes	yes	yes	yes
Vulkan API trace	yes	yes	no	no
Direct3D12, Direct3D11, DXR, & PIX APIs	-	yes	-	-
OpenXR	-	yes	-	-
OptiX	7.1+	7.1+	-	-
Bidirectional correlation of API and GPU workload	yes	yes	yes	yes
Identify GPU idle and sparse usage	yes	yes	yes	yes
Multi-GPU Graphics trace	OpenGL and Vulkan	Direct3D12, OpenGL and Vulkan	-	-
Trace graphics resource migration between VRAM and System Memory	-	yes	-	-
Ready for big data
Fast GUI capable of visualizing in excess of 10 million events on laptops	yes	yes	yes	yes
Additional command line collection tool	yes	no	no	no
NV-Docker container support	yes	-	-	-
NVIDIA GPU Cloud support	yes	-	-	-
Minimum user privilege level	user	administrator	root	root

* On Intel Haswell and newer CPU architecture
** Only with OS runtime trace enabled. Some syscalls such as handcrafted assembly may be missed. Backtraces may only appear if time threasholds are exceeded.

---

Archives

Access older versions of Nsight Systems in the Gameworks Download Center.
View older version release notes in the Nsight System’s documentation archive.

---

Resources

Nsight Systems Documentation

You can also learn about installing & using the NVIDIA Tools Extension API (NVTX) here.

Nsight Tools Tutorial Center

Access the latest resources to get started with Nsight Systems.

---

Access Self-Paced Training

Nsight Systems Documentation

Get hands on training for Nsight Systems with self-paced online courses from the NVIDIA Deep Learning Institute.

See more courses on Accelerated Computing for Developers.

---

Tutorial Sessions

---

Video Series

Learn about using Nsight Systems for CUDA Development in the CUDA Developer Tools tutorial series.

---

Support

To provide feedback, request additional features, or report support issues, please use the Developer Forums .