REPORT WU-SPEEDS HERE !

MAGIC Quantum M...

Joined: 18 Jan 05

Posts: 1890

Credit: 1414011128

RAC: 1182402

> JAHMAGIC, > > Just

3 Mar 2005 11:01:48 UTC

Message 5816 in response to message 5810

(moderation:

)

> JAHMAGIC,
>
> Just wondering, do you work at Fermilab?
>

Yes, but I have been retired for about 8 years now and that was over 10 years ago.

Back when Fermilab's founder, the late physicist Robert Wilson was still alive.

(he is still there at the complex)

It is quite a place......6,800-acre complex that looks like a National Park!

Of course with particle accelerators under the ground.

http://www.fnal.gov/

I imagine since you are at Illinois Institute of Technology that you may have been to Batavia.

Jordan Wilberding

Joined: 19 Feb 05

Posts: 162

Credit: 715454

RAC: 0

> I imagine since you are at

3 Mar 2005 14:23:15 UTC

Message 5817 in response to message 5816

(moderation:

)

> I imagine since you are at Illinois Institute of Technology that you may have
> been to Batavia.

I actually have a co-op there every other semester. It definately is a very cool place. I just hope they get the new linear accelerator they are vying for, so they don't die out after the LHC gets completed at CERN.

such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell

zulfy26

Joined: 26 Feb 05

Posts: 1

Credit: 61954

RAC: 0

P4 1.6 Ghz 15 hours per WU

3 Mar 2005 22:33:15 UTC

Message 5818

(moderation:

)

P4 1.6 Ghz 15 hours per WU with between 75-95 claimed credit.

Juerschi

Joined: 4 Jan 05

Posts: 62

Credit: 31245

RAC: 0

AMD Athlon XP-M 2600+ 6h45min

6 Mar 2005 22:18:41 UTC

Message 5819

(moderation:

)

AMD Athlon XP-M 2600+ 6h45min per WU

Steve-O

Joined: 9 Feb 05

Posts: 1

Credit: 5081

RAC: 0

HP Pavillion a220n Athlon XP

7 Mar 2005 1:40:50 UTC

Message 5820 in response to message 5819

(moderation:

)

HP Pavillion a220n
Athlon XP 2600+ at 2.083 MHz, locked.
Asus A7N8X-VM 512MB SDRAM

wu times=6:20 to 6:45.

For only 2.083, I think it's a great time.

G Thomas Wilson

Joined: 5 Mar 05

Posts: 21

Credit: 2311944

RAC: 0

I just started einstein@home

7 Mar 2005 3:40:12 UTC

Message 5821

(moderation:

)

I just started einstein@home this weekend as seti@home seems to be searching for electricians as opposed to aliens :D

I haven't used boinc nearly long enough yet to compile good averages.
Here's what I have come up with so far...

AMD Athlon XP 2600+ ( @ 2.25 GHz ) Linux
seti@home: ~02h:30m
einstein@home: ~10h:20m

-----------------------------------
AMD Sempron 2200+ (1.5 GHz) Linux
seti@home: ~04h:40m
einstein@home: ~15h:10m

-----------------------------------
AMD Athlon 2800+ (2.0 GHz) Linux
seti@home: ~03h:40m
einstein@home: ~10h:50m

-----------------------------------
Intel Pentium 4 (1.80 GHz) Linux
seti@home ~05h:30m
einstein@home ~17h:05m

-----------------------------------
Intel Pentium 4 (1.60 GHz) Linux
seti@home ~07h:00m
einstein@home ~19h:24m

-----------------------------------
AMD Athlon K6/2 (700 MHz) Linux
seti@home ~10h:05m
einstein@home ~24h:10m

-----------------------------------
Intel Pentium 4 (1.40 GHz) Linux
seti@home ~07h:45m
einstein@home ~23h:05m

-----------------------------------
Intel Pentium 4 (1.70 Ghz) Linux
seti@home ~06h:05m
einstein@home ~18h:10m

-----------------------------------
Intel Pentium III (450 MHz) Linux
seti@home ~15h:25m
einstein@home ~48h:03m (estimated)

-----------------------------------
Intel Pentium III Coppermine (1.0 GHz) Linux
seti@home ~09h:05m
einstein@home ~24h:16m (estimated)

-----------------------------------
AMD K6 3D (450 MHz) Linux
seti@home ~10h:42m (One very small WU)
einstein@home ~102h:19m (estimated)

-----------------------------------
Intel Pentium III (866 MHz)
seti@home ~09h:15m
einstein@home ~39h:50m (estimated)

RandyC

Joined: 18 Jan 05

Posts: 6617

Credit: 111139797

RAC: 0

> I just started

7 Mar 2005 13:20:53 UTC

Message 5822 in response to message 5821

(moderation:

)

> I just started einstein@home this weekend as seti@home seems to be searching
> for electricians as opposed to aliens :D
>
> I haven't used boinc nearly long enough yet to compile good averages.
> Here's what I have come up with so far...
>
I see most of your systems are running under Linux. Linux does Einstein VERY poorly. Check out this thread as a way to possibly boost your throughput:

http://einsteinathome.org/node/188222

Seti Classic Final Total: 11446 WU.

Jarmo Kahila

Joined: 20 Feb 05

Posts: 1

Credit: 89756379

RAC: 0

Same AMD XP 2400+ computer

7 Mar 2005 16:52:01 UTC

Message 5823

(moderation:

)

Same AMD XP 2400+ computer takes 22000 seconds / packet with windows 2000 and 38000 seconds with linux. Even worst thing is that einstein gives zero credit at half of all ready made linux packets (not overclocked). So einstein program has been removed from all of my linux computers until...

FalconFly

Joined: 16 Feb 05

Posts: 191

Credit: 15650710

RAC: 0

AMD Athlon64 3700+

7 Mar 2005 21:49:03 UTC

Message 5824 in response to message 5823

(moderation:

)

AMD Athlon64 3700+ (Win2000)
FPU : 2267.72 million ops/sec
INT: 6284.15 million ops/sec
Avg : 18.9k sec

AMD Athlon64 3200+ (Linux)
FPU : 2134.64 million ops/sec
INT : 4893.04 million ops/sec
Avg : 38.7k sec

AMD Athlon64 3000+ (Linux)
FPU : 2138 million ops/sec
INT : 4878.45 million ops/sec
Avg : 37.7k sec

AMD AthlonXP 3000+ (Win2000)
FPU : 2002.54 million ops/sec
INT : 4820.02 million ops/sec
Avg : 21.6k sec

AMD AthlonXP 3000+ (Linux)
FPU : 2257.9 million ops/sec
INT : 4770.85 million ops/sec
Avg : 35.8k sec

AMD AthlonXP 2800+ (Linux)
FPU : 2191.91 million ops/sec
INT : 4340.49 million ops/sec
Avg : 36.4k sec

Dual AMD AthlonMP 2800+ (Linux)
FPU : 2252.69 million ops/sec
INT : 4458.22 million ops/sec
Avg : 35.8k sec

AMD AthlonXP 2700+ (Linux)
FPU : 2269.82 million ops/sec
INT : 4813.92 million ops/sec
Avg : 35.7k sec

AMD AthlonXP 2600+ (Linux)
FPU : 2181.68 million ops/sec
INT : 4538.67 million ops/sec
Avg : 37.1k sec

AMD AthlonXP 2500+ (Linux)
FPU : 1933.56 million ops/sec
INT : 3829.77 million ops/sec
Avg : 41.2k sec

AMD AthlonXP 2400+ (Linux)
FPU : 2102.69 million ops/sec
INT : 4149.54 million ops/sec
Avg : 38.2k sec

AMD AthlonXP 2200+ (Linux)
FPU : 1886.7 million ops/sec
INT : 3986.49 million ops/sec
Avg : 42.9k sec

AMD AthlonXP 2100+ (Linux)
FPU : 1820.19 million ops/sec
INT : 3648.92 million ops/sec
Avg : 45.4k sec

Dual AMD AthlonMP 2000+ (Linux)
FPU : 1759.1 million ops/sec
INT : 3483.2 million ops/sec
Avg : 48k sec

AMD AthlonXP-M 1800+ (WinXP)
FPU : 1389.65 million ops/sec
INT : 3383.78 million ops/sec
Avg : 30.9k sec
--------------------------------------------
From those numbers it becomes clear that :
- the E@H Linux Client is absolutely terrible in terms of performance (adding to the 0 Credits/Invalidation bug)
- it loves raw clock rates
- it does not take advantage of larger (512k+) L2 Caches

(Performance numbers based on still relatively few WorkUnits, therefor subject to certain variation)

nsteussy

Joined: 11 Nov 04

Posts: 4

Credit: 31984

RAC: 0

AMD Barton 2500+, 32bit,

8 Mar 2005 0:24:52 UTC

Message 5825

(moderation:

)

AMD Barton 2500+, 32bit, Fedora Core 3, average einstein wu over 21 results, 42k seconds.

AMD64 2600+, 64bit, Fedora Core 3 x86_64, average enstein uw over 11 results 78k seconds.

Something is majorly wrong with their code running on AMD64 in 64bit linux.

Duke out

REPORT WU-SPEEDS HERE !

Forums › Cruncher's Corner

Comment viewing options

Forums › Cruncher's Corner