Jump to content
Sign in to follow this  
smily03

Max CPU frequency

Recommended Posts

So I decided to try to push the 8 cores a bit and see how they'd perform. So I installed "stress" from the Ubuntu repositories, and ran it for 8 CPUs. top ramped up to 100% utilization on all 8 cores nicely, and the load averages popped up just like I'd hope. I let this run for about 30 minutes to see how things would pan out.

 

top - 01:42:18 up 2 days, 23:08,  5 users,  load average: 7.99, 7.65, 7.59

Tasks: 134 total,  13 running, 121 sleeping,   0 stopped,   0 zombie
%Cpu0  : 99.7 us,  0.0 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.0 hi,  0.3 si,  0.0 st
%Cpu1  :100.0 us,  0.0 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu2  : 98.7 us,  1.3 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu3  :100.0 us,  0.0 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu4  :100.0 us,  0.0 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu5  :100.0 us,  0.0 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu6  :100.0 us,  0.0 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu7  :100.0 us,  0.0 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
KiB Mem:    894656 total,   154320 used,   740336 free,     8436 buffers
KiB Swap:        0 total,        0 used,        0 free.    88528 cached Mem

  PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
 2661 root      20   0    1640     72      0 R  90.3  0.0   3:54.44 stress
 2665 root      20   0    1640     72      0 R  80.7  0.0   3:48.43 stress
 2664 root      20   0    1640     72      0 R  73.0  0.0   3:45.85 stress
 2666 root      20   0    1640     72      0 R  71.7  0.0   3:47.95 stress
 2663 root      20   0    1640     72      0 R  66.9  0.0   3:48.04 stress
 2660 root      20   0    1640     72      0 R  63.7  0.0   3:41.22 stress
 2662 root      20   0    1640     72      0 R  62.7  0.0   3:45.22 stress
 2659 root      20   0    1640     72      0 R  61.4  0.0   3:43.03 stress
 2568 root      20   0    4504   1208    824 R   1.6  0.1   0:32.56 top

 

So I decided to check cpufreq-info to see how the cores were running, and noticed that they were all capped at 1.61GHz, even though they're supposedly capable of 2.02GHz.

 

analyzing CPU 7:
  driver: sunxi-iks
  CPUs which run at the same hardware frequency: 0 1 2 3 4 5 6 7
  CPUs which need to have their frequency coordinated by software: 0 1 2 3 4 5 6 7
  maximum transition latency: 2.00 ms.
  hardware limits: 480 MHz - 1.80 GHz
  available frequency steps: 2.02 GHz, 1.80 GHz, 1.61 GHz, 1.20 GHz, 1.13 GHz, 1.01 GHz, 912 MHz, 864 MHz, 720 MHz, 600 MHz, 480 MHz
  available cpufreq governors: interactive, conservative, ondemand, userspace, powersave, performance
  current policy: frequency should be within 480 MHz and 1.61 GHz.
                  The governor "interactive" may decide which speed to use
                  within this range.
  current CPU frequency is 1.61 GHz (asserted by call to hardware).
  cpufreq stats: 2.02 GHz:0.00%, 1.80 GHz:0.02%, 1.61 GHz:0.67%, 1.20 GHz:0.19%, 1.13 GHz:0.00%, 1.01 GHz:0.00%, 912 MHz:0.00%, 864 MHz:0.02%, 720 MHz:0.00%, 600 MHz:0.00%, 480 MHz:99.10%  (2101)

 

 

root@whiteboard:/home/linaro# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_max_freq
1608000

 

Out of curiosity, I checked the ACPI thermal zone, and it was hovering right around 68C

 

root@whiteboard:/# cat /sys/class/thermal/thermal_zone0/temp
68

 

So I killed the stress test, and it dropped down pretty quickly. But I also noticed that as the internal temp dropped, the max_freq increased. Even still, when it cooled down, the max was still 1.8GHz, which cpufreq-info said was a hardware limit (even though it was capable of 2.02GHz?)

 

So, long story short -- if you plan on really pushing this board, you'll need to add some extra cooling to the CPU, otherwise it'll dumb itself down on you.

 

 

 

Share this post


Link to post
Share on other sites
Guest son_of_panoflex

we had a discussion on the quality of this crappy board.  it is the stupid idea of powering via USB and not barrel connector for DC input.  they have good CPU which throttles itself and expecting to push this board beyond what you have going you may as well just make it a doorstop LOL. 

 

i find this $10 USD board only good for simple things.  it was ok to learn from because so little works out of the box and you have to push your patience trying to get results plus you dont mind taking a hammer to it if it makes you mad.

Share this post


Link to post
Share on other sites
Guest
You are commenting as a guest. If you have an account, please sign in.
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
Sign in to follow this  

×