This forum uses cookies
This forum makes use of cookies to store your login information if you are registered, and your last visit if you are not. Cookies are small text documents stored on your computer; the cookies set by this forum can only be used on this website and pose no security risk. Cookies on this forum also track the specific topics you have read and when you last read them. Please confirm whether you accept or reject these cookies being set.

A cookie will be stored in your browser regardless of choice to prevent you being asked this question again. You will be able to change your cookie settings at any time using the link in the footer.

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Tinkerboard beats Raspberry PI4 in HPL Linpack!
#1
Compiled HPL Linpack 2.3 test on both boards (Tinkerboard v1.2 and RPi 4B [4GB])

Results:
Raspberry Pi4: 6 Gflops
Tinkerboard:    7.14 Gflops!

Both tests ran on all 4 CPU cores in double precision



Tinker output:
===================================================

T/V                N    NB     P     Q               Time                 Gflops
--------------------------------------------------------------------------------
WR11C2R4       14208   192     2     2             267.66             7.1449e+00
HPL_pdgesv() start time Fri Jul 12 22:59:54 2019

HPL_pdgesv() end time   Fri Jul 12 23:04:22 2019

--------------------------------------------------------------------------------
||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)=   1.64513478e-03 ...... PASSED
===================================================
Reply
#2
Looks like the pi 4 can be overclocked to 2ghz so that may make a difference.
Another win for the tinkerboard is that it has an audio DAC onboard for the 3.5mm audio output, the PI 4 just uses PWM and a few diode and capacitors to produce audio, so you really need a usb or gpio DAC for proper audio on the PI 4.

I guess the best part of the pi 4 is usb 3, so booting from a usb 3 sd card reader should be a lot faster.

Also open office seems to run a lot better on the pi 4, using tinker OS openoffice CALC seems to really struggle

Someone has done benchmarks on PI 4 and Tinkerboard

https://www.phoronix.com/vr.php?view=28057


average scores are (higher is better)

tinkerboard 16.38
and
pi 4 4gb 19.4
Reply
#3
Tinkerboard still rules with android though,
Pi has always been weak on android, with no official image.

Pi4 now has fast usb3 though, that makes a huge difference.
Reply
#4
(07-13-2019, 07:48 AM)Z80 Wrote: Another win for the tinkerboard is that it has an audio DAC onboard for the 3.5mm audio output, the PI 4  just uses PWM and a few diode and capacitors to produce audio, so you really need a usb or gpio  DAC for proper audio on the PI 4.

Then I wonder what's the normal price for a good DAC. If u see what I mean.
And also, the audio output was much better on my RPI 1 model B with Librelec than on my Tinker actually, no comparison. It's a fact.
Reply
#5
what about pi4 Vs tinkerboardS boot time given S has emmc.

if pi4B+ came out with emmc would this improve things?
...working my way through the 1000+ ways things don't work to find the one that does...
Reply
#6
It's my feeling that Raspberry (family) has a huge background and documentation to make things working with easy.
So, it might not be a matter if TB still have some better performance, but after one will purchase it might give some negative impression. In particular for the support.
Light blue words might be a link. Have you try to click on them? Big Grin
Reply
#7
Star 
Now, how will they behave together? (I mean Pi4 + Tinker)

So, for compatibility, I compiled the same version of openmpi-4.0.1 on both boards, then recompiled HPL benchmark against new openmpi, joined both boards in small MPI cluster via 1Gbps lan, assigned IPs, and fired up the benchmark... So, when working in pair the total performance is... 

10 Gflops 

(similar to a 10 years old AMD laptop I had ages ago.. Smile. Still there is an improvement vs. each board working separately...

Here I launched xhpl with mpirun using hostfile, listing all the hosts (chiefly just two: pi4 and tinker, repeated as many times as the number of CPUs each of therm has, obviously /etc/hosts must list their IPs to be able to find each other)

pi@tinker:~/HPL$ /opt/openmpi4/bin/mpirun -np 8 -hostfile hosts.txt xhpl
================================================================================
HPLinpack 2.3  --  High-Performance Linpack benchmark  --   December 2, 2018
Written by A. Petitet and R. Clint Whaley,  Innovative Computing Laboratory, UTK
Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
Modified by Julien Langou, University of Colorado Denver
================================================================================

An explanation of the input/output parameters follows:
T/V    : Wall time / encoded variant.
N      : The order of the coefficient matrix A.
NB     : The partitioning blocking factor.
P      : The number of process rows.
Q      : The number of process columns.
Time   : Time in seconds to solve the linear system.
Gflops : Rate of execution for solving the linear system.

The following parameter values will be used:

N      :   20352 
NB     :     192 
PMAP   : Row-major process mapping
P      :       2 
Q      :       4 
PFACT  :   Right 
NBMIN  :       4 
NDIV   :       2 
RFACT  :   Crout 
BCAST  :  1ringM 
DEPTH  :       1 
SWAP   : Mix (threshold = 64)
L1     : transposed form
U      : transposed form
EQUIL  : yes
ALIGN  : 8 double precision words

--------------------------------------------------------------------------------

- The matrix A is randomly generated for each test.
- The following scaled residual check will be computed:
      ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
- The relative machine precision (eps) is taken to be               1.110223e-16
- Computational tests pass if scaled residuals are less than                16.0

================================================================================
T/V                N    NB     P     Q               Time                 Gflops
--------------------------------------------------------------------------------
WR11C2R4       20352   192     2     4             515.90             1.0895e+01
HPL_pdgesv() start time Mon Jul 15 19:23:35 2019

HPL_pdgesv() end time   Mon Jul 15 19:32:11 2019

--------------------------------------------------------------------------------
||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)=   9.58007276e-04 ...... PASSED
================================================================================

Finished      1 tests with the following results:
              1 tests completed and passed residual checks,
              0 tests completed and failed residual checks,
              0 tests skipped because of illegal input values.
--------------------------------------------------------------------------------

End of Tests.
================================================================================
Reply
#8
You should post the two results side by side.
Light blue words might be a link. Have you try to click on them? Big Grin
Reply


Forum Jump:


Users browsing this thread: 2 Guest(s)