3 ms·

My app slows 10%, on an OS that slows 5%, running in a virtualization host that slows 10%. The numbers add up.

• Florin_Andrei 9y ago
  Well, it's not a slowdown really. The high performance of the physical CPU was achieved through means that compromise security. The top of the speed range was bullshit, basically. Now that the top is being lopped off, you experience the true speed of that CPU while operating in a secure manner.
  • ghughes 9y ago
    I hear Intel's PR department is hiring, if you're interested.
  • fao_ 9y ago
    > Now that the top is being lopped off, you experience the true speed of that CPU while operating in a secure manner. So the proper execution is slower than the secure execution, right? Ergo, it is 'slower'.
• fvdessen 9y ago
  Do they ? If your usage is 50% userspace, 50% kernel, then having each slice slow by 10% does not result in a 20% slowdown.
  • gpm 9y ago
    They do... because they are measured in a weird way. If the usage is split 50/50 kernel/userspace. And the userspace changes slow the app down 10%. Then the user space changes slowed the userspace code down 20%. Likewise if the kernel changes slow the app down 10%, they slowed the kernel code down by 20%. So the total impact of the changes is 20% in your example.
    • fvdessen 9y ago
      That is indeed a weird way to measure and report. Without the associated userspace/kernel usage ratio those numbers are meaningless.
      • gpm 9y ago
        They aren't meaningless, they are measuring exactly what you care about. How much more computing power do you need to achieve the job.