Menu

Individual skill

intel-vtune-amd-uprof

Use when profiling with Intel VTune or AMD uProf for hotspots, top-down pipeline stalls, memory-bound analysis, or roofline data. Not for raw perf stat counters: use linux-perf.

.devin/skills/intel-vtune-amd-uprof/SKILL.md

Read pinned source

Source commit b0e8ce89a19fac880251dc3ea1babfeb4503a4fe

What was checked

Matches the observed repository HEAD
Yes
Scanner rules are current
Yes
Repository scan complete
No
Skill scan complete
No
Observed behavior
Instructions only
Last checked
Nov 4, 2026

These are observations at the recorded commit, not a guarantee of safety. Review the source before running it.

Install this skill

Choose an agent in your Stack or copy a command below. Commands fetch the pinned commit and copy only this skill directory. Existing destinations are refused. Copying a command does not install anything.

An install command is unavailable: current rules, complete repository and skill scans, an eligible behavior level and a permitted license are required.

Repository context

Repository stars (not skill ratings)
55
Repository last updated
Sep 28, 2026
License
Apache-2.0

Observed agent integration: No specific integration evidence. Standard SKILL.md installation does not imply verified compatibility.

Files and excerpt

File metadata and permitted, redacted excerpts come from the recorded scan. Source links are pinned to that commit.

Browse files
Read excerpt

# Intel VTune and AMD uProf

VTune Profiler (a free download, standalone or inside the oneAPI toolkits) and AMD uProf (5.3, 2026-06-17) are the vendor profilers that turn PMU samples into named bottleneck categories. Use them after `linux-perf` has named a hot function and the question has become why that function is slow: front end, bad speculation, memory, or core execution.

## Contract

| Field | Bound contract |
|---|---|
| Trigger | A hot function is known and the user asks which pipeline stage or memory level limits it, wants DRAM bandwidth utilization, wants a roofline chart, or asks how to run VTune or uProf. |
| Authority | Reversible local: writes only result directories (`-result-dir` or `--output-dir` targets) and report files in the working directory; rollback is deleting them. Driver installation for hardware sampling is proposed to the user, never performed. No remote mutation. |
| Side effect | Result directories on disk. Hardware event collection loads the vendor sampling driver or uses `perf_event_open` under the same paranoid rules as perf. |
| Done | The report names the dominant top-down category for the hot function with its percentage, the memory level or bandwidth figure behind it when memory-bound, and the measured versus peak bandwidth or the roofline position when asked. |

## Inputs