i need about 8 hours for each work unit
which surprises me (on windows platform it just takes 2 and a half hour) and the old(classic) setiathome was faster too

my datas:

- gentoo linux (2.6.14-gentoo-r3)
- AMD Athlon(tm) 64 Processor 3000+
- setiathome: 4.18 x86_64

i have no idea where is the problem, i gave seti 100% cpu power
and the workunits are also not bigger than at the old classic setiathome

anybody knows why it is so slowly and how to improve/optimize that?

every help is apprecitated

regards
schorfi

You could use an optimized setiathome app.

Try this one ---> http://naparst.name/seti.htm (Pentium4 SSE2 should work)
if it doesn't work you could try one of my builds.
AthlonXP_SSE_FFTW3

---

Join BOINC United now!

i need about 8 hours for each work unit
which surprises me (on windows platform it just takes 2 and a half hour) and the old(classic) setiathome was faster too

Some people had a simliar problem. Some systems had a problem with the power management. Can you please post the output from "cat /proc/cpuinfo" here.

Please look also in this thread : http://setiathome.berkeley.edu/forum_thread.php?id=18354

---

i disabled the cool n quite feature (shout down the prog)
but to make sure:
cat /proc/cpuinfo:
processor : 0
vendor_id : AuthenticAMD
cpu family : 15
model : 4
model name : AMD Athlon(tm) 64 Processor 3000+
stepping : 8
cpu MHz : 2002.652
cache size : 512 KB
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext lm 3dnowext 3dnow
bogomips : 4014.50
TLB size : 1024 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp


moreover the messure speeds of the CPUs on windows and here are nearly equal
will try another seti build

if my memory is correct, the current setiathome is not self-compiled o_O
so not really optimize for my system

i disabled the cool n quite feature (shout down the prog)
but to make sure:
cat /proc/cpuinfo:
processor : 0
vendor_id : AuthenticAMD
cpu family : 15
model : 4
model name : AMD Athlon(tm) 64 Processor 3000+
stepping : 8
cpu MHz : 2002.652
cache size : 512 KB
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext lm 3dnowext 3dnow
bogomips : 4014.50
TLB size : 1024 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp


moreover the messure speeds of the CPUs on windows and here are nearly equal
will try another seti build

if my memory is correct, the current setiathome is not self-compiled o_O
so not really optimize for my system

You should really try my suggestion.
You'll see that'll work really fast.

---

Join BOINC United now!

moreover the messure speeds of the CPUs on windows and here are nearly equal
will try another seti build

if my memory is correct, the current setiathome is not self-compiled o_O
so not really optimize for my system

I really have to re-iterate what Crunch3r said - use Harold's P4-SSE2 build on your AMD64. After a lot of trial and error, i found it to be the fastest on my AMD64 4000+ and Opteron 146.

This is one of those rare occasions when compiling your own on Gentoo just isn't worth it and, probably, counter-productive. That would certainly be the case if you were to use the Berkeley source. Harold's source contains some fairly radical changes to greatly improve the way the calculations are done.

Crunch3r's XP build is by far the fastest on my XP3200+s. ;)

---

Be lucky

Neil

just compile your own custom client for your rig and the speed will be sexy. I compiled a client along with a ck kernel and i jumped from 30 average to 35 average credit. a 64 bit cpu and linux are a little shaky so a custom compile is the way to go!

---

just compile your own custom client for your rig and the speed will be sexy. I compiled a client along with a ck kernel and i jumped from 30 average to 35 average credit. a 64 bit cpu and linux are a little shaky so a custom compile is the way to go!

Compiling the BOINC client is one thing - although even that needs some thought to get the best results. See Ned slider's site for information on that. However, the topic of this conversation is the Seti app. That is an entirely different kettle of fish.

To be totally frank and brutal, your RAC and TC is laughable and could be improved drastically by using one of Crunch3r's compiled apps. I suggest you read a thread fully and make sure you understand it before posting such blatant garbage in contradiction of far more experienced people.

---

Be lucky

Neil