I'm up to #59 in the team rankings
Trying hard to break into that top 20
I need more computers......
EDIT: 57
Now ----- #53
now
50 .......
Are you running two instances of the F@H smp client per machine, or just one instance per machine? The F@H smp core is designed to run optimally on 4 cores. With a 8 core machine, I believe you could run two instances of the client. If you are only running one instance, I'm betting that running two instances would come close to doubling your output to about 6000 points per day per machine.
well, i forgot to point that out; i am allready running two instances:
Cpu0 : 0.0%us, 2.1%sy, 35.0%ni, 62.5%id, 0.0%wa, 0.0%hi, 0.3%si, 0.0%st
Cpu1 : 0.0%us, 2.3%sy, 42.2%ni, 55.1%id, 0.0%wa, 0.0%hi, 0.4%si, 0.0%st
Cpu2 : 0.0%us, 2.7%sy, 66.7%ni, 30.2%id, 0.0%wa, 0.0%hi, 0.4%si, 0.0%st
Cpu3 : 0.0%us, 2.5%sy, 57.3%ni, 39.9%id, 0.0%wa, 0.0%hi, 0.4%si, 0.0%st
Cpu4 : 0.0%us, 2.4%sy, 45.8%ni, 51.5%id, 0.0%wa, 0.0%hi, 0.4%si, 0.0%st
Cpu5 : 0.0%us, 2.3%sy, 42.7%ni, 54.7%id, 0.0%wa, 0.0%hi, 0.3%si, 0.0%st
Cpu6 : 0.0%us, 2.4%sy, 52.8%ni, 44.5%id, 0.0%wa, 0.0%hi, 0.4%si, 0.0%st
Cpu7 : 0.0%us, 2.5%sy, 57.9%ni, 39.3%id, 0.0%wa, 0.0%hi, 0.4%si, 0.0%st
Mem: 8179984k total, 1114948k used, 7065036k free, 149172k buffers
Swap: 8193128k total, 0k used, 8193128k free, 454452k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
570 root 34 19 100m 65m 2548 S 103 0.8 56:35.40 FahCore_a1.exe
32631 root 34 19 100m 65m 2556 S 96 0.8 520:28.26 FahCore_a1.exe
32632 root 34 19 91776 55m 2352 S 93 0.7 474:45.46 FahCore_a1.exe
571 root 34 19 91464 55m 2324 S 78 0.7 51:51.42 FahCore_a1.exe
32633 root 34 19 85804 49m 1936 S 61 0.6 323:41.44 FahCore_a1.exe
572 root 34 19 85688 49m 1960 S 61 0.6 35:24.34 FahCore_a1.exe
32634 root 34 19 84780 48m 2208 S 39 0.6 216:53.71 FahCore_a1.exe
573 root 34 19 84732 48m 2224 S 39 0.6 23:23.29 FahCore_a1.exe
Even if the computer center machines are only temporary, it all helps.:) Besides, you are actually doing the computer center a favor by stress testing their new machines. If there are any intermittent problems lurking in a new machine, Folding @ Home will find them. If you are only running one instance of the client, then running two instances would stress test it even more.
:>
]]>I have two lower-spec machines @home not running fah all the time and two equal machines at the local computer centre; non-mod debian-64 kernel, two quad-core Xeons /w 8Gb RAM
i wont be able to let them work for "us" on long term, they have just been installed and nobody cares
Finkregh;
Thanks for posting your info on the Quad core Xeons. I wish I had a couple of those! Let me make sure I have this straight. At the computer center, you have two machines with 8 cores each, for a total of 16 cores? I noticed for the last 4 days you have averaged 6000 points per day. That would be 3000 points per day per machine if I have this straight. Are you running two instances of the F@H smp client per machine, or just one instance per machine? The F@H smp core is designed to run optimally on 4 cores. With a 8 core machine, I believe you could run two instances of the client. If you are only running one instance, I'm betting that running two instances would come close to doubling your output to about 6000 points per day per machine. In a terminal window, run top and see what percentage of CPU usage each instance of FahCore_a1.exe is using. Here is a snippet of my top:
PID USER PR NI VIRT RES SHR S % CPU % MEM TIME+ COMMAND
6646 don 39 19 94576 55m 2368 S 57 2.9 234:02.88 FahCore_a1.exe
6645 don 39 19 103m 65m 2520 S 51 3.5 215:18.46 FahCore_a1.exe
6647 don 39 19 88420 49m 2024 S 39 2.6 163:48.88 FahCore_a1.exe
6648 don 39 19 87784 48m 2228 S 24 2.6 100:07.20 FahCore_a1.exe
The way I understand it, the F@H smp client ALWAYS starts 4 threads, irregardless of how many cores you have. So, even though I only have a Dual Core machine, the F@H smp client starts four threads, so each core is running 2 threads. Not as efficient as four cores, which would run one thread per core. Unless I'm mistaken, on a 8 core machine, one instance of the client still only starts 4 threads. To achieve one thread per core, you would need to run two instances of the F@H smp client.
Even if the computer center machines are only temporary, it all helps.:) Besides, you are actually doing the computer center a favor by stress testing their new machines. If there are any intermittent problems lurking in a new machine, Folding @ Home will find them. If you are only running one instance of the client, then running two instances would stress test it even more.
Pudge
]]>well, yes. i visit the forums from time to time
I have two lower-spec machines @home not running fah all the time and two equal machines at the local computer centre; non-mod debian-64 kernel, two quad-core Xeons /w 8Gb RAM
i wont be able to let them work for "us" on long term, they have just been installed and nobody cares
# uname -a
Linux XXX 2.6.18-5-amd64 #1 SMP Mon Dec 3 04:55:28 UTC 2007 x86_64 GNU/Linux
# cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Intel(R) Xeon(R) CPU E5320 @ 1.86GHz
stepping : 7
cpu MHz : 1866.809
cache size : 4096 KB
physical id : 0
siblings : 4
core id : 0
cpu cores : 4
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall nx lm constant_tsc pni monitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm
bogomips : 3736.63
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management:
[...]
processor : 7
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Intel(R) Xeon(R) CPU E5320 @ 1.86GHz
stepping : 7
cpu MHz : 1866.809
cache size : 4096 KB
physical id : 1
siblings : 4
core id : 3
cpu cores : 4
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall nx lm constant_tsc pni monitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm
bogomips : 3733.81
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management:
# free -m
total used free shared buffers cached
Mem: 7988 1087 6900 0 145 444
édit: i use the following switches: ./fah6 -local -forceasm -smp -advmethods -verbosity 9
]]>Pudge wrote:Sekre wrote:Should really get x64 up and running :S
Didn't know there was that much of a difference actually.x64 single core doesn't make that much difference. x86 Dual core or Quad core makes a BIG difference using the SMP client.
Pudge
A C2D would be dual x64 so that will be good then yes ?
If you have Arch 64 bit installed as the Operating System, and you are using the SMP Folding @ Home client, then that would be good. You will probably get around 900 points per day on that dual core running the SMP client.
Pudge
]]>Sekre wrote:Should really get x64 up and running :S
Didn't know there was that much of a difference actually.x64 single core doesn't make that much difference. x86 Dual core or Quad core makes a BIG difference using the SMP client.
Pudge
A C2D would be dual x64 so that will be good then yes ?
]]>We'll get there in no time because Arch folders are using a freaking awesome OS
]]>Team Ranking (incl. aggregate) 276 of 90466
If we keep it up, maybe we can pass PCLinuxOS team
]]>Should really get x64 up and running :S
Didn't know there was that much of a difference actually.
x64 single core doesn't make that much difference. x86 Dual core or Quad core makes a BIG difference using the SMP client.
Pudge
]]>Now......... to attract more members to fold for the team
]]>I don't think I will beat ya crouse, but I shall do my best hehe =P
I wonder if there might be a friendly underground computer center nearby... >_>
I'm catching ya pudge.... I'm now at # 84 lol lol
Bring it on Dude!:) Fold harder, fold faster.
Pudge
]]>