What is a simple safe way to test the speed and verify the results?

I think my computer might be running slow.
HP Z-Workstation

I am running Kubuntu 24.04.2 LTS (Noble Numbat).
The CPU is 12th Gen Intel(R) Core™ i7-12700K with 32gig of ram no gpu

I still have a little time on my warranty.

George

Hey @Fred here’s some ideas we got discussing this on the Discord server. If you want to see the conversation itself on the Discord then you can go here.

These steps are just a guide with some ideas, not a guaranteed fix. You also probably wont need to do everything.

Basic testing that wont require stress testing.

  • Compare with a clean environment
    • Boot a Kubuntu 24.04 live USB (or Fedora KDE/EndeavourOS KDE).
    • Use it for 10–15 minutes; note any difference in snappiness.

Why? If the live session feels fine, the slowdown lives in your installed system. If not, it’s likely hardware.

  • Audit background services & autostart apps
    • In your installed Kubuntu, open System Monitor or htop → sort by CPU and Memory.
    • Disable unused services in System Settings → Startup & Shutdown → Background Services.
    • Uninstall heavy applications you don’t need.

Why? Kubuntu ships a sane default set, but any distro can get bogged down by extras over time.

  • Check the graphics stack
    • Make sure you’re on the latest mesa and intel-media-driver (sudo apt update && sudo apt upgrade).
    • In System Settings → Display & Monitor → Compositor, briefly disable desktop effects and see if the UI perks up.

Why? The Intel iGPU is plenty fast, but driver hiccups or fancy effects can still cause jank.


If the answer isnt found there then might need to do some stress testing.

  • Stress-test the hardware
    • Install OCCT (or stress-ng / phoronix-test-suite / s-tui / Hardinfo2).
    • Run CPU, RAM and power tests for at least 10–15 minutes each.
    • Watch for errors, crashes, or scores that are way below published results for the i7-12700K.

Why? This quickly confirms whether the silicon itself is healthy.

  • Watch thermals and throttling
    • While the stress test runs, open sensors, Psensor, or btop and display temps & clocks.
    • Compare idle vs. load temps. Check if clocks drop when temps rise.

Why? High temps or clock drops mean the cooler, thermal paste, or airflow needs attention—and you can still claim on warranty.

  • Check disk and I/O bottlenecks
    • During normal use, run iotop -oPa (or enable iowait in btop).
    • Look for a process that’s constantly hitting the disk.
    • Run smartctl -a /dev/nvme0 (or /dev/sdX) to verify drive health.

Why? Storage stalls often feel like “the whole system is slow”.

  • Validate memory health & speed
    • Boot memtest86+ overnight.
    • Enter BIOS and confirm the RAM is running at its rated speed (e.g., 3200 MHz, not 2400 MHz).

Why? Bad or down-clocked RAM can both crash and drag a system’s pace.

  • Escalate under warranty if needed
    • Save screenshots of failed OCCT runs, memtest errors, SMART logs, or thermal graphs.
    • Open an HP ticket while coverage remains.

Why? Hard evidence speeds up an RMA and tries to avoid the “works fine in our lab” response.

1 Like

Thank you for the comprehensive response.