How fast does this run on your PC? - Page 9

Evil_Bert · #**121** 18th December 2011, 10:52 PM

Code:

real	0m5.676s
user	0m5.591s
sys	0m0.057s

AMD Phenom II X4 945 at stock speed (3.0 GHz), running F14 x64, with bogomips 6027.24 (apparently).

Version of gmp-devel is gmp-devel-4.3.1-7.fc14.x86_64.

pinkfloyd · #**122** 19th December 2011, 12:33 PM

grep -e model -e cache -e MHz -e bogo /proc/cpuinfo | head -5

model : 2
model name : AMD Athlon(tm) 7750 Dual-Core Processor
cpu MHz : 2700.247
cache size : 512 KB
bogomips : 5400.49

Code:

real	0m6.659s
user	0m6.516s
sys	0m0.098s

flyingfsck · #**123** 19th December 2011, 01:25 PM

On my Lenovo S10 netbook:

time ./mersenne 43112609 > M46

real 1m54.901s
user 1m53.934s
sys 0m0.664s

The poor little thing almost had a heart attack and is still wheezing...

Bigger numbers are better right?

rclark · #**124** 19th December 2011, 02:59 PM

My Fedora 16 box :

real 0m24.025s
user 0m23.715s
sys 0m0.182s

Gödel · #**125** 19th December 2011, 09:16 PM

Quote:

Originally Posted by RupertPupkin

I finally figured out the source of my problem: cpu frequency scaling is horribly broken for my processor in F16. For some infuriating reason the Fedora developers decided to eliminate cpu frequency scaling as an optional service as it had been in earlier versions of Fedora; it is now compiled directly into the kernels that ship with F16.

So I had to build a custom kernel(!) with cpu frequency scaling disabled. My cli apps now run fast again, including the mersenne program:

Code:

$ time ./mersenne 43112609 > M46

real    0m9.076s
user    0m8.481s
sys     0m0.073s

Yay! (But boo on the Fedora devs for their lame-brained decision.

)

That's a pretty terrible bug. If it's still there in f17 development you should let them know.

In most bioses you can usually disable frequency scaling, but this basic stuff should work. I couldn't manage with the cpu on my laptop at constant maxspeed, but luckily ondemand frequency scaling works on intel core2 duos.

ah7013 · #**126** 31st December 2011, 10:19 AM

Thought I would try it on my new laptop

Code:

[andrew@localhost ~]$ time ./mersenne 43112609 > M46

real    0m3.045s
user    0m2.996s
sys     0m0.043s

[andrew@localhost ~]$ grep -e model -e cache -e MHz -e bogo /proc/cpuinfo | head -5
model           : 42
model name      : Intel(R) Core(TM) i7-2640M CPU @ 2.80GHz
cpu MHz         : 800.000
cache size      : 4096 KB
bogomips        : 5586.71

[andrew@localhost ~]$ rpm -q lib64gmp-devel
lib64gmp-devel-5.0.1-3.mga1

markus39 · #**127** 31st December 2011, 10:56 AM

Code:

$ time ./mersenne 43112609 > M46

real	0m12.859s
user	0m12.769s
sys	0m0.054s

4 year old Lenovo T61p, 4GB ram with Fedora 16 64bit

EDIT 04/02/2012
My T61p died two weaks ago...

On my new Lenovo W520, 16GB with Fedora 16 64bit
model : 42
model name : Intel(R) Core(TM) i7-2860QM CPU @ 2.50GHz
cpu MHz : 800.000
cache size : 8192 KB
bogomips : 4983.46

real 0m5.579s
user 0m5.531s
sys 0m0.024s

JamesNZ · #**128** 6th January 2012, 10:45 PM

Turion X2 dual-core 2.1GHz, running F16:

Code:

$ time ./mersenne 43112609 > M46

real	0m10.222s
user	0m10.057s
sys	0m0.109s

RupertPupkin · #**129** 12th January 2012, 05:45 PM

Using the koji gmp-devel-5.0.2-3.fc17 package in F16 makes it much faster than before (it took over 9 seconds using gmp-devel-4.3.2-4.fc16):

Code:

$ time ./mersenne 43112609 > M46

real    0m5.697s
user    0m5.219s
sys     0m0.103s

sea · #**130** 13th January 2012, 10:06 AM

On the Notebook: (gmp-devel-4.3.2.-4)

Code:

real	0m56.735s
user	0m56.389s
sys	0m0.199s

On the Laptop: (gmp as above)

Code:

real	0m16.949s
user	0m16.515s
sys	0m0.046s

VernDog · #**131** 13th January 2012, 09:39 PM

HP Pavilion Slimline s5310y:

Installed: gmp-devel.x86_64 1:4.3.2-4.fc16

Quote:

real 0m5.820s
user 0m5.699s
sys 0m0.061s

Quote:

model : 6
model name : AMD Athlon(tm) II X2 250 Processor
cpu MHz : 800.000
cache size : 1024 KB
bogomips : 6027.24

smr54 · #**132** 15th January 2012, 03:49 AM

ArchLinux on a UX31e Zenbook (SSD drive) i5 @1.70 (According to cpuinfo) 4GB RAM

Code:

real 0m3.935s
user 0m3.893s
sys 0m0.033s

F16 on same machine (no adjustments made for CPU scaling mentioned earlier)

Code:

real 0m7.315s
user 0m7.247s
sys 0m0.038s

smr54 · #**133** 11th May 2012, 01:19 PM

And, on a brand new toy--Clevo laptop with i7 processor and 16GB of RAM

real 0m0.113.s
user 0m0.107s
sys 0m0.005s

EDIT--that might be wrong, can't reproduce it, or possibly my aging eyes misread.

real 0m3.080s
user 0m3.024s
sys 0m0.030s

I'm getting that or similar after running a few times.

sphinxx · #**134** 29th December 2012, 06:46 AM

Intel(R) Core(TM) i5-2430M CPU @ 2.40GHz

[root@aman aman]# time ./mesc 43112609 > M46

real 0m3.695s
user 0m3.615s
sys 0m0.049s

Fenrin · #**135** 29th December 2012, 07:17 PM

AMD Phenom II X2 @ 3.00 GHz (2 Core), Fedora 18 beta x86_64

$ time ./mersenne 43112609 > M46

real 0m3.396s
user 0m3.323s
sys 0m0.060s