Pi 3 vs Pi 2 @1.1ghz ?

[DavidTheDolphin]

Bonjour,

J'ai lu des comparatifs sur les différences de perfs entre le Pi 2 et le Pi 3 à leurs fréquences nominales hors j'ai lu aussi qu'il n'était pas permis d'OC le Pi3 alors que mes Pi 2 tournent à 1.1 ghz...

Quelqu'un aurait des infos sur les différences entr'un Pi 2 OC et un Pi 3 SVP ?

Merci,
A+
David

[totoharibo]

le PI2 a 2 cœurs, le RPI3 4.

Ça dépend de ce que tu fais et si tes applis utilisent tous les cœurs.
Si le WiFi et le Bluetooth intégré t'intéressent alors ...

Si tu as besoin de plus de puissance regarde les autres monocartes avec un processeur ARM. Attention toutefois au soutien logiciel sur ces cartes.

[general alcazar]

le rpi2 est un quadcore.

Code: Select all

pi@raspberrypi2:~ $ cat /proc/cpuinfo 
processor	: 0
model name	: ARMv7 Processor rev 5 (v7l)
BogoMIPS	: 57.60
Features	: half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm 
CPU implementer	: 0x41
CPU architecture: 7
CPU variant	: 0x0
CPU part	: 0xc07
CPU revision	: 5

processor	: 1
model name	: ARMv7 Processor rev 5 (v7l)
BogoMIPS	: 57.60
Features	: half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm 
CPU implementer	: 0x41
CPU architecture: 7
CPU variant	: 0x0
CPU part	: 0xc07
CPU revision	: 5

processor	: 2
model name	: ARMv7 Processor rev 5 (v7l)
BogoMIPS	: 57.60
Features	: half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm 
CPU implementer	: 0x41
CPU architecture: 7
CPU variant	: 0x0
CPU part	: 0xc07
CPU revision	: 5

processor	: 3
model name	: ARMv7 Processor rev 5 (v7l)
BogoMIPS	: 57.60
Features	: half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm 
CPU implementer	: 0x41
CPU architecture: 7
CPU variant	: 0x0
CPU part	: 0xc07
CPU revision	: 5

Hardware	: BCM2709
Revision	: a01041
Serial		: 000000003fdd6ef

[totoharibo]

Mes excuses : j'ai utilisé ma mémoire.
"j'ai la mémoire qui flanche" ...<"
Je confondais avec le B+ qui n'a qu'un cœur.

Ma remarque reste valable : un quadricœur pour les softs qui n'utilisent qu'un cœur ça ne le fait pas.

[JumpZero]

C'est surtout un nouveau SoC! Pour les perfs voir ici https://www.raspberrypi.org/magpi/raspb ... enchmarks/

[Rive]

Bonjour,

J'ai lu des comparatifs sur les différences de perfs entre le Pi 2 et le Pi 3 à leurs fréquences nominales hors j'ai lu aussi qu'il n'était pas permis d'OC le Pi3 alors que mes Pi 2 tournent à 1.1 ghz...

Quelqu'un aurait des infos sur les différences entr'un Pi 2 OC et un Pi 3 SVP ?

Merci,
A+
David

You can overclock the pi3. It will easily run at 1.3 GHz.

The pi3 has twice the performance of the pi2.

[DavidTheDolphin]

Thank you for your answer but this kind of OC will void waranty

[DavidTheDolphin]

C'est surtout un nouveau SoC! Pour les perfs voir ici https://www.raspberrypi.org/magpi/raspb ... enchmarks/

Merci mais je doute qu'il s'agisse d'un comparatif entr'un Pi3 et un Pi2 OC à 1.1ghz.

[Rive]

Thank you for your answer but this kind of OC will void waranty

Since when? That is news to me. (but even if that is true, it is just $35 in any event....which is a tank of petrol/gas)

[DavidTheDolphin]

Pi 3 doesn't officially support overclocking.

[Rive]

Pi 3 doesn't officially support overclocking.

Show me where overclocking the pi3 voids warranty. Just because there are no raspi-config overclock presets, doesn't mean overclocking isn't supported, or that doing so voids manufacturer warranty.

[JumpZero]

Thank you for your answer but this kind of OC will void waranty

No, overclocking doesn't void the waranty
Overvolting does https://www.raspberrypi.org/blog/introd ... -for-free/
It's an old post, but it's Eben, and you can read a lot on the forums about this and afaik it's still the same.
I agree with Rive, if there is no oc presets in raspi-config menu it doesn't mean it isn't supported.
--
Jmp0

[DavidTheDolphin]

https://www.element14.com/community/doc ... tions-faqs

Can the Pi 3 be overclocked?
Overclocking the Raspberry Pi 3 is not recommended and you may void your warranty attempting to do so.

[Rive]

https://www.element14.com/community/doc ... tions-faqs

Can the Pi 3 be overclocked?
Overclocking the Raspberry Pi 3 is not recommended and you may void your warranty attempting to do so.

Wow. that is like super vague, and as such I think you may be getting confused. Notice the wording "MAY", not "wILL".

The pi3 costs the same as a tank of petrol/fuel. Even If I were to void my Pi3 warranty, I consider it disposable. I would sooner trash it, and buy another before I would jump through hoops filing a warranty claim, or paying for shipping on it.


In the end though, it is your call. Even stock, the Pi3 is twice as fast/powerful than the Pi2, and for the same price, AND it can be easily overclocked to even further improve performance and speed for those who like that kind of stuff, or not.

Mine is overclocked, and you can see the benchmarks, and performance on it here:

I did the Linpack Bench, SysBench, SdBench, and Cpuburn-a53
See: viewtopic.php?p=952371#p952371

[DavidTheDolphin]

Potentialy my boss will pay a lot of Pi 3 so the waranty is important !

My Pi 2 are OC @1.1ghz, the object of my post is to know if a Pi 3 (with his waranty !) has much more perf than Pi 2 @1.1ghz

[Rive]

Potentialy my boss will pay a lot of Pi 3 so the waranty is important !

My Pi 2 are OC @1.1ghz, the object of my post is to know if a Pi 3 (with his waranty !) has much more perf than Pi 2 @1.1ghz

The answer, again, would still be yes.

[DavidTheDolphin]

Do you have a benchmark please ?

[Rive]

Do you have a benchmark please ?

I have two: Sysbench with validated run, and Linpack bench :

SYSBENCH

Run sysbench on your overclocked Pi2, and compare your total time. Then you will answer you own question regarding performance of Stock Pi3 versus the Pi2 @ 1.1 GHz.

For comparison, the Stock Pi2 runs the validated sysbench in 191 Seconds, wheras the stock Pi3 runs it in 119 seconds.

http://www.raspberrypistarterkits.com/2 ... rk-review/



Sysbench

Code: Select all

sudo apt-get install sysbench

run Sysbench with validation

Code: Select all

sysbench --num-threads=4 --test=cpu --cpu-max-prime=20000 --validate run

Results of Sysbench on Stock Pi3: total time: 119.9360s

Code: Select all

pi@raspberrypi:~ $ sysbench --num-threads=4 --test=cpu --cpu-max-prime=20000 --validate run
sysbench 0.4.12:  multi-threaded system evaluation benchmark

Running the test with following options:
Number of threads: 4
Additional request validation enabled.


Doing CPU performance benchmark

Threads started!
Done.

Maximum prime number checked in CPU test: 20000


Test execution summary:
    total time:                          119.9360s
    total number of events:              10000
    total time taken by event execution: 479.6257
    per-request statistics:
         min:                                 47.69ms
         avg:                                 47.96ms
         max:                                168.58ms
         approx.  95 percentile:              48.26ms

Threads fairness:
    events (avg/stddev):           2500.0000/8.60
    execution time (avg/stddev):   119.9064/0.02

LINPACK

Run the linpack on your overclocked Pi2, then compare PASSED Gflops. Then you will answer you own question regarding performance of Stock Pi3 versus the Pi2 @ 1.1 GHz (if the benchmark fails/crashes, then your overclock is faulty/unstable).

For comparison, a Stock Pi2 is 156 Mflops (or 0.156 Gflops), and a Pi2 at 1.0 GHz, is 176 Mflops (or 0.176 Gflops)....Compare this to a Pi3 stock at 6 Gflops.

Pi2 Linpack Bench:
http://www.roylongbottom.org.uk/Raspber ... hmarks.htm



Linpack Benchmark:

Code: Select all

sudo apt-get install libmpich-dev
wget http://web.eece.maine.edu/~vweaver/junk/pi3_hpl.tar.gz
tar -xvzf pi3_hpl.tar.gz 
chmod +x xhpl

run Linpack Benchmark:

Code: Select all

./xhpl

For other benchmarks you can run for further comparison, see:
viewtopic.php?p=952371#p952371


Results for Pi3 (Stock) PASSED Linpack benchmark (N=8000): Gflops 6.16

Code: Select all

pi@raspberrypi:~ $ ./xhpl
================================================================================
HPLinpack 2.1  --  High-Performance Linpack benchmark  --   October 26, 2012
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      :    8000 
NB     :     256 
PMAP   : Row-major process mapping
P      :       1 
Q      :       1 
PFACT  :    Left 
NBMIN  :       2 
NDIV   :       2 
RFACT  :   Right 
BCAST  :   2ring 
DEPTH  :       0 
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
--------------------------------------------------------------------------------
WR02R2L2        8000   256     1     1              55.37              6.166e+00
HPL_pdgesv() start time Sat Apr 23 15:14:17 2016

HPL_pdgesv() end time   Sat Apr 23 15:15:12 2016

--------------------------------------------------------------------------------
||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)=        0.0025941 ...... 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.

[DavidTheDolphin]

Thanx

--> Pi 2 @1.1ghz vs Pi 3 stock <--

Sysbench : 177.8687s vs 119.9360s

Code: Select all

sysbench --num-threads=4 --test=cpu --cpu-max-prime=20000 --validate run

sysbench 0.4.12:  multi-threaded system evaluation benchmark

Running the test with following options:
Number of threads: 4
Additional request validation enabled.


Doing CPU performance benchmark

Threads started!
Done.

Maximum prime number checked in CPU test: 20000


Test execution summary:
    total time:                          177.8687s
    total number of events:              10000
    total time taken by event execution: 711.3230
    per-request statistics:
         min:                                 70.55ms
         avg:                                 71.13ms
         max:                                201.39ms
         approx.  95 percentile:              71.13ms

Threads fairness:
    events (avg/stddev):           2500.0000/6.04
    execution time (avg/stddev):   177.8308/0.02

Linmark benchmark (N=8000) : ~ vs 6.16 Gflops

How long may I wait ???

[Rive]

Linpack benchmark (N=8000) : ~ vs 6.16 Gflops

How long may I wait ???

It only takes 3-5 min to run. if you freeze up, then you crashed, and may have an unstable overclock, overheated, or have insufficient power/voltage.

The way I run linpack is I keep an eye on the clock (bottom right corner of screen), and if the seconds stop turning over...means crash.

To recover from a crash/lockup on the pi2/pi3, reboot from keyboard with 'Alt'+'SysRQ'+'b'

Here is a video: https://youtu.be/lm97OHawd-8

[DavidTheDolphin]

OK :/

I'm in an SSH console so i have no clock to look at... But I've been waiting for 90 minutes now !

[Rive]

OK :/

I'm in an SSH console so i have no clock to look at... But I've been waiting for 90 minutes now !

Sounds like a crash (which basically means your pi2 spectacularly failed the benchmark, and is in a nutshell incapable/or borderline incapable of running the benchmark as configured....usually as a result of unstable overclocking, voltage/power issues, or cooling issues)...See if you can do anything on the pi2, if it is locked up/frozen, you may need to reboot/restart.

FYI, on a stable, properly cooled, properly powered Pi, the Linpack Benchmark only runs for a few minutes or so (as demonstrated in the video above).

[DavidTheDolphin]

OK, new try with a little more CPU voltage (1.35V instead of 1.3) :

First attempt --> segmentation fault (core dumped)
Second attempt --> freeze again

Higher voltage will set waranty bit...

In "normal" conditions @1.1ghz/1.3V I add no freeze until this bench

I've got a 2A Aukru power adapter

[Rive]

OK, new try with a little more CPU voltage (1.35V instead of 1.3) :

First attempt --> segmentation fault (core dumped)
Second attempt --> freeze again

Higher voltage will set waranty bit...

In "normal" conditions @1.1ghz/1.3V I add no freeze until this bench

Instead of trying to raise voltage, lower the arm_freq...that is probably what is wrong. 1.1 GHZ is likely unstable.

[DavidTheDolphin]

OK and your results show that Pi 2 @1.0ghz is far away behind Pi 3 stock !

I've got my answer, thank you