Please note that this blog has been moved.
Now it has its own domain: mynixworld.info 🙂
If you want to read the latest version of this article (recommended) please click here and I open the page for you.
As my Compaq 6820s laptop’s AC/DC died (or maybe only the laptop’s DC jack, I haven’t investigated yet the issue) I had been forced to move on and to put at work an old friend, a HP dc7900eC/E8400 convertible minitower.
The laptop has a Corsair Force 3 SSD 90GB hard disk and I wished desperately to use it again.
If I was a Windows user I would be in that situation when you have to start all from the scratch: reinstall OS, apps, drivers, …, move data from the old system to the new system.
Well, the good news is that I am not a Windows user. I am a Linux user and I am proud about this. So all I had to do was to move the SSD from the broken laptop to my old-friend dc7900 desktop. It booted without any problem, so few minutes later I was able to continue my usual work.
But I was not willing to talk about that, what I intended to write was about the performance I got by using the old-friend dc7900 desktop PC.
After upgrading my laptop CPU I have been using an Intel T9300 that gave my a lot of satisfaction using that laptop. On the dc7900eC desktop I have an Intel E8400 cpu that is 0.5 GHz faster. The performance difference between them can be seen here. They have the same number of cores and threads but the FSB of the desktop is 1333MHz while the FSB of the laptop is only 800MHz. Also the bus/core ratio is 9 for the desktop and 12.5 for the laptop. The rest of their specifications are quite similar.
If you have the speed of the northbridge bus used to carry data between CPU and memory hub (aka FSB), the bus/core ratio and the number of core/threads of the cpu then you can do the math and you will get the cpu speed. For instance:
- Intel T9300 (2 cores, 2 threads per core, FSB 800MHz, bus/core ratio is 12.5)
- speed(cpu) = ratio*fsb/threads = 12.5*800MHz/(2*2) = 2500MHz
- Intel E8400 (2 cores, 2 threads per core, FSB 1333MHz, bus/core ratio is 9)
- speed(cpu) = ratio*fsb/threads = 9*1333MHz/(2*2) = 2999.25MHz
Ok, so I got 0.5MHz more, so what?
If after the installation of the SSD disk in my laptop I was impressed to find that the SSD boosts my laptop boot from 20″ to something like 12 seconds (which means 40% faster boot), now I am in a situation where my desktop system boots in about 6 seconds. This is, by the way, 50% faster than ‘yesterday’. However, this is 70% faster than the day I used for the first time my laptop (which now is broken, if I haven’t told you yet).
I should mention that my system is a Linux 3.5.0 #2 SMP x86_64 Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz GenuineIntel GNU/Linux and I use Xfce as a desktop environment. The boot is timed up to the moment that the desktop is fully loaded (with auto-login in my case) and the user can starts using it.
Do my system really boots in just only 6 second? Let’s see:
Note: in the image above see the vertical dotted line which shows where the boot really finished. The boot is considered as finished when the Xfce window manager have started (e.q. xfwm4 process). That point correspond with the moment when the desktop was loaded and ready to accept the user requests.