What is NVIDIA Nsight Systems?
NVIDIA Nsight Systems is a profiling and tracing tool for CUDA C++ programs. Instead of a single live reading like nvidia-smi, it records a full run and lays CPU activity, GPU kernels, and memory transfers out together on one timeline, so you can see where time actually went instead of guessing from an aggregate number.
Nobody picks a GPU because it's pleasant to write code for; teams reach for one because ordinary hardware isn't fast enough for the job, which makes nearly every serious CUDA program performance-sensitive by nature. Nsight Systems exists for that pressure: its timeline turns "something in here is slow" into an inspectable trace, and it's the natural next step once a kernel needs checking against the roofline model to see whether it's actually compute-bound.
Aquanode instances ship with a standard NVIDIA driver and CUDA toolchain, so Nsight Systems traces a rented marketplace GPU the same way it would a workstation you own.
Building on GPUs? Aquanode runs the workload.
Deploy on H100, H200, B200, A100 and MI300X across a multi-provider marketplace, without racking your own hardware or committing to one cloud's spec sheet.
See also
nvidia-smi
nvidia-smi is the command line tool for querying and managing NVIDIA GPUs, built on the NVML management library. What it reports, what it can change, and why its text output isn't a stable interface.
Roofline Model
The roofline model plots a kernel's arithmetic intensity against two hardware ceilings, memory bandwidth and arithmetic bandwidth, to show at a glance whether it's compute-bound or memory-bound. Where it came from and why GPUs need it.
Compute-bound
A compute-bound kernel is limited by arithmetic throughput rather than memory bandwidth. When LLM inference hits this regime, and a back-of-the-envelope estimate for the batch size it takes to get there.
