Gimme Atom timings!

nuggetprime · 2011-02-17, 19:46

Yes,I know Atoms suck somewhat for number crunching stuff,but I'm still interested in timings as I'm planning to extend my crunchforce by something low-power which will run 24/7.
I'm not that interested in 10MDigit mersenne stuff,but instead somewhat smaller candidates using LLR.
So if you have a netbook/nettop/something else with an Intel Atom CPU(any model,please tell me which one it's got) and can waste a bit of CPU time: get LLR 3.8.5 for your operating system from jpenne.free.fr,unzip it and run:
Linux:

Code:

./llr -d -q325*2^100001-1

Windows
[open command prompt and go to directory where LLR binary is]

Code:

llr -d -q325*2^100001-1

and post the output here.

Thanks
Nugget

Harvey563 · 2011-02-18, 03:09

steven@steven-eee:~/Desktop$ ./llr -d -q325*2^100001-1
Starting Lucas Lehmer Riesel prime test of 325*2^100001-1
Using FFT length 6K
325*2^100001-1 is not prime. LLR Res64: 1784445AA7974710 Time : 111.009 sec.
steven@steven-eee:~/Desktop$ cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 28
model name : Intel(R) Atom(TM) CPU N270 @ 1.60GHz
stepping : 2
cpu MHz : 800.000
cache size : 512 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 1
apicid : 0
initial apicid : 0
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de tsc msr pae mce cx8 apic mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx constant_tsc arch_perfmon pebs bts aperfmperf pni dtes64 monitor ds_cpl est tm2 ssse3 xtpr pdcm movbe lahf_lm
bogomips : 3192.32
clflush size : 64
cache_alignment : 64
address sizes : 32 bits physical, 32 bits virtual
power management:

processor : 1
vendor_id : GenuineIntel
cpu family : 6
model : 28
model name : Intel(R) Atom(TM) CPU N270 @ 1.60GHz
stepping : 2
cpu MHz : 800.000
cache size : 512 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 1
apicid : 1
initial apicid : 1
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de tsc msr pae mce cx8 apic mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx constant_tsc arch_perfmon pebs bts aperfmperf pni dtes64 monitor ds_cpl est tm2 ssse3 xtpr pdcm movbe lahf_lm
bogomips : 3191.95
clflush size : 64
cache_alignment : 64
address sizes : 32 bits physical, 32 bits virtual
power management:

steven@steven-eee:~/Desktop$

nuggetprime · 2011-02-18, 12:06

That's about 8 times slower than on my P4,so Atom can't really impress for primality testing.
Does anybody know some other low-power system (<20 W) that has better performance for this kind of work?

fivemack · 2011-02-18, 18:36

Quote:

Originally Posted by nuggetprime

Yes,I know Atoms suck somewhat for number crunching stuff,but I'm still interested in timings as I'm planning to extend my crunchforce by something low-power which will run 24/7.

Basically, the fact that computers have lots of things in them other than the CPU means that optimum performance-per-watt for pure-CPU jobs comes from the lowest-end version of fast computers ... an underclocked i5/2500 with just enough memory and the OS on a USB stick is probably as good performance-per-watt as you can get.

imwithid · 2011-02-18, 22:07

I was hoping someone would post such info. Now I'm curious to see that compared to more recent versions (e.g. D510). AFAIK, the N270 is the bottom of the line for the Atom series. My understanding is that it would be comparable to a Centrino of similar rating (in this case, 1.6 Ghz for the D510).

ldesnogu · 2011-02-21, 08:28

All currently available Atom cores share the same micro-architecture. So I'd expect single thread performance to be very similar on all of them (of course, availability of 64-bit could make a difference, but anyway Atom is a low performer across all the range).

2011-02-17, 19:46	#1
nuggetprime Mar 2007 Austria 2×151 Posts	Gimme Atom timings! Yes,I know Atoms suck somewhat for number crunching stuff,but I'm still interested in timings as I'm planning to extend my crunchforce by something low-power which will run 24/7. I'm not that interested in 10MDigit mersenne stuff,but instead somewhat smaller candidates using LLR. So if you have a netbook/nettop/something else with an Intel Atom CPU(any model,please tell me which one it's got) and can waste a bit of CPU time: get LLR 3.8.5 for your operating system from jpenne.free.fr,unzip it and run: Linux: Code: ./llr -d -q3252^100001-1 Windows [open command prompt and go to directory where LLR binary is] Code: llr -d -q3252^100001-1 and post the output here. Thanks Nugget

2011-02-18, 03:09	#2
Harvey563 Apr 2004 3·61 Posts	N270 @ 1.6 Ghz steven@steven-eee:~/Desktop$ ./llr -d -q3252^100001-1 Starting Lucas Lehmer Riesel prime test of 3252^100001-1 Using FFT length 6K 325*2^100001-1 is not prime. LLR Res64: 1784445AA7974710 Time : 111.009 sec. steven@steven-eee:~/Desktop$ cat /proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 28 model name : Intel(R) Atom(TM) CPU N270 @ 1.60GHz stepping : 2 cpu MHz : 800.000 cache size : 512 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 1 apicid : 0 initial apicid : 0 fdiv_bug : no hlt_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 10 wp : yes flags : fpu vme de tsc msr pae mce cx8 apic mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx constant_tsc arch_perfmon pebs bts aperfmperf pni dtes64 monitor ds_cpl est tm2 ssse3 xtpr pdcm movbe lahf_lm bogomips : 3192.32 clflush size : 64 cache_alignment : 64 address sizes : 32 bits physical, 32 bits virtual power management: processor : 1 vendor_id : GenuineIntel cpu family : 6 model : 28 model name : Intel(R) Atom(TM) CPU N270 @ 1.60GHz stepping : 2 cpu MHz : 800.000 cache size : 512 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 1 apicid : 1 initial apicid : 1 fdiv_bug : no hlt_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 10 wp : yes flags : fpu vme de tsc msr pae mce cx8 apic mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx constant_tsc arch_perfmon pebs bts aperfmperf pni dtes64 monitor ds_cpl est tm2 ssse3 xtpr pdcm movbe lahf_lm bogomips : 3191.95 clflush size : 64 cache_alignment : 64 address sizes : 32 bits physical, 32 bits virtual power management: steven@steven-eee:~/Desktop$

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Intel Atom revisited	hj47	Hardware	15	2010-07-08 20:19
Intel Atom CPU	storm5510	Hardware	15	2009-08-21 02:35
New proggy and timings	axn	Operation Billion Digits	1	2009-02-06 16:14
Worth running on Atom?	db597	Hardware	8	2009-01-28 09:45
321 LLR timings	paulunderwood	3*2^n-1 Search	14	2008-04-17 22:27

2011-02-18, 12:06	#3
nuggetprime Mar 2007 Austria 2·151 Posts	That's about 8 times slower than on my P4,so Atom can't really impress for primality testing. Does anybody know some other low-power system (<20 W) that has better performance for this kind of work?

2011-02-18, 22:07	#5
imwithid Apr 2009 Venice, Chased by Jaws 1010111₂ Posts	I was hoping someone would post such info. Now I'm curious to see that compared to more recent versions (e.g. D510). AFAIK, the N270 is the bottom of the line for the Atom series. My understanding is that it would be comparable to a Centrino of similar rating (in this case, 1.6 Ghz for the D510).

2011-02-21, 08:28	#6
ldesnogu Jan 2008 France 2·5²·11 Posts	All currently available Atom cores share the same micro-architecture. So I'd expect single thread performance to be very similar on all of them (of course, availability of 64-bit could make a difference, but anyway Atom is a low performer across all the range).