Skip to content

Latest commit

 

History

History
43 lines (35 loc) · 3.66 KB

File metadata and controls

43 lines (35 loc) · 3.66 KB

Competitive review

An overview of existing Windows CPU-monitoring tools — what they do and what's missing for our scenario: "notice that a single core has been under load for a while, and name the culprit."

Comparison table

# Tool Per-core load Tray informativeness Sustained-load alerts Names the culprit Price
1 Windows Task Manager (tray icon) ✗ (overall %) minimal — a single green meter ✗ (manual only) built-in
2 XMeters (bars in taskbar) high — per-core bars right in the taskbar free (personal)
3 SysStatsTray medium — dynamic icons, color by load free
4 Process Lasso (ProBalance) medium — graph in tray — reacts to CPU-hogging processes freemium
5 HWiNFO (per-core sensors) medium — several values as separate tray icons — thresholds on any sensor free
6 Libre Hardware Monitor medium — chosen sensors as numbers in tray open source
7 Core Temp per-core (temperature) medium — per-core temps in tray yes (overheating) free
8 TrafficMonitor ✗ (overall CPU) high — floating window / taskbar embed open source
9 tray-monitor high — graph icons in tray top processes on click open source
10 Performance Monitor / perfmon (built-in) (Processor Information counters) — Data Collector Sets + event-log alerts built-in

Conclusions

  1. Per-core visualization is offered by XMeters, HWiNFO and Libre Hardware Monitor — but none of them can alert specifically on a single core being under sustained load.
  2. Alerts exist in Process Lasso (by process, not by core), HWiNFO (by sensor thresholds, without a duration condition and without a culprit) and perfmon (powerful, but an admin tool: no UI notifications, configured through Data Collector Sets).
  3. The culprit is named only by Process Lasso, but its model is "a process exceeded its overall CPU quota" — not "core #5 has been busy at 100% for a minute."
  4. The scenario "one stuck core at low overall load" is covered by nobody: on a 16-core CPU a hung single-threaded process is ~6% overall and triggers none of the tools above.

CorePulse's niche is to combine three things that already exist separately: a per-core tray icon (like XMeters) + alerts with a duration condition (like ProBalance/perfmon) + naming the responsible process (like Process Lasso), in one lightweight app that needs no administrator rights.

Sources