I thought this deserves a new thread...

The first Barcelona on SETI
The second Barcelona on SETI

Andy.

In that case, I'd better put my reply in here too!

The first Barcelona on SETI

Ran 2 WUs (rather slowly), tried and failed to optimise, retired hurt for the weekend? Not very impressive for an 8-core.

The second Barcelona on SETI

At least it's crunched everything it's been given. Slightly faster than my E5320s on stock, but hardly significant. I'll keep an eye on it, and if it runs steadily I might do a comparative plot.

I've looked at the "claimed credit" and tried to find them on my puters. I found no matching ones on the AMD64 X2 6000, or 5200, but did find these on the AMD64 X2 4800 running stock app:

0.006445 5.10.20 604694483 154038229 2 Sep 2007 16:35:00 UTC 3 Sep 2007 10:44:25 UTC Over Success Done 13,396.97 63.98 63.98

It's claimed credit "63.98" matches this ONE of his/hers:

606821891 155019947 6 Sep 2007 12:41:15 UTC 6 Sep 2007 20:20:06 UTC Over Success Done 14,631.83 63.98 63.98

it's AR was 0.00666

I've also found these of mine (AMD64 X2 4800):
1.490466 5.10.20 604351720 153883107 2 Sep 2007 20:20:18 UTC 3 Sep 2007 16:11:54 UTC Over Success Done 3,920.83 19.31 19.31
1.492136 5.10.20 604351838 153883150 2 Sep 2007 20:20:18 UTC 4 Sep 2007 0:21:38 UTC Over Success Done 4,242.48 19.31 19.31
1.492136 5.10.20 604351938 153883212 2 Sep 2007 20:20:18 UTC 3 Sep 2007 18:40:10 UTC Over Success Done 4,244.48 19.31 19.31
1.492136 5.10.20 604351944 153883197 2 Sep 2007 20:20:18 UTC 4 Sep 2007 0:21:38 UTC Over Success Done 4,214.72 19.31 19.31
1.490466 5.10.20 604351946 153883203 2 Sep 2007 20:20:18 UTC 4 Sep 2007 0:21:38 UTC Over Success Done 4,222.41 19.31 19.3
1.492136 5.10.20 604352128 153883295 2 Sep 2007 20:20:18 UTC 3 Sep 2007 16:11:54 UTC Over Success Done 3,918.84 19.31 19.31
1.492136 5.10.20 604352166 153883316 2 Sep 2007 20:20:18 UTC 3 Sep 2007 12:02:29 UTC Over Success Done 4,311.67 19.31 19.31
1.490466 5.10.20 604352194 153883320 2 Sep 2007 20:20:18 UTC 3 Sep 2007 14:36:29 UTC Over Success Done 4,126.61 19.31 19.31
1.492136 5.10.20 604352196 153883327 2 Sep 2007 20:20:18 UTC 3 Sep 2007 18:40:10 UTC Over Success Done 4,259.48 19.31 19.31
1.490466 5.10.20 604352294 153883362 2 Sep 2007 20:20:18 UTC 4 Sep 2007 0:21:38 UTC Over Success Done 4,268.80 19.31 19.31
1.492136 5.10.20 604352296 153883369 2 Sep 2007 20:20:18 UTC 4 Sep 2007 0:21:38 UTC Over Success Done 4,258.63 19.31 19.31
1.490466 5.10.20 604352298 153883368 2 Sep 2007 20:20:18 UTC 4 Sep 2007 0:21:38 UTC Over Success Done 4,163.84 19.31 19.31
1.490466 5.10.20 604352368 153883411 2 Sep 2007 20:20:18 UTC 3 Sep 2007 18:40:10 UTC Over Success Done 4,429.14 19.31 19.31
1.490466 5.10.20 604352370 153883396 2 Sep 2007 20:20:18 UTC 3 Sep 2007 18:40:10 UTC Over Success Done 4,199.81 19.31 19.31
1.492136 5.10.20 604352412 153883431 2 Sep 2007 20:20:18 UTC 4 Sep 2007 0:21:38 UTC Over Success Done 4,197.69 19.31 19.31
which claimed the same, but the AR on his/hers was 1.488811 for these two:

607409480 3773162 8 Sep 2007 22:32:11 UTC 9 Sep 2007 0:58:41 UTC Over Success Done 4,398.34 19.31 19.31
607409478 3773162 8 Sep 2007 22:32:11 UTC 9 Sep 2007 0:58:41 UTC Over Success Done 4,882.91 19.31 19.31

NOTE: I've so far found no exact matches, and all shown wus are from stock 5.27

Yes, looks like some tweaking going on, and both using stock apps too, so still some room for improvement, but doesn't look to have dropped a bombshell yet.

I just did a shout over in Boincstats, having found them there too...

Boincstats uses the same computer ids per project, but the id for the combined results are internal, though fortunately the page also accepted the host cpid to be able to find them:

http://www.boincstats.com/stats/boinc_host_graph.php?pr=bo&id=2791692
and
http://www.boincstats.com/stats/boinc_host_graph.php?pr=bo&id=2796867

Shall follow with interest...

OK, I've set up a chart to plot the second one against the stock data I've got saved for my E5320s. So far, all the results are within the scatter range for the Xeons, but with 9 WUs that doesn't prove much.

Setting up the sheet for reference, I noticed that he/she is running Enterprise Server in 512MB of RAM - not a terribly well-balanced combination, even for a test machine?

One thing I can't understand...

System 1 = 4gb ram, 1691.28 mfps, 3617.42 mis

System 2 = 512mb ram, 1929.78 mis, 4248.44 mfps

Could this be down to the o/s?

---

Yes, need more results yet... doesn't help that the project is 'scrooging' right now!

512Mb should be enough if all the machine is doing is crunching. Anyone running this with such a new processor will only be evaluating and testing it, particularly as it is only officially released today, so will still be a while before more people get their mitts on it.

Sirius B, the benchmark in boinc doesn't really test RAM.

Sirius B, the benchmark in boinc doesn't really test RAM.

Sorry Astro, I only put the ram in to highlight the difference.

I thought the mis & mfps is down to the cpu?

---

There is also This system to draw statistics from.

/Anton

There is also This system to draw statistics from.

/Anton

Ah yes, missed it because of the different product code. It is actually the first one to be registered!

After searching for "Quad-Core AMD" instead of "Processor 8347", Barcelona hosts on SETI currently stand at 3:

<host>
<id>3768873</id>
<userid>8239893</userid>
<total_credit>1949.346608</total_credit>
<expavg_credit>148.047934</expavg_credit>
<expavg_time>1189366764.582640</expavg_time>
<p_vendor>AuthenticAMD</p_vendor>
<p_model>Quad-Core AMD Opteron(tm) Processor 8350</p_model>
<os_name>Microsoft Windows XP</os_name>
<os_version>Professional Edition, Service Pack 2, (05.01.2600.00)</os_version>
<create_time>1188986661</create_time>
<rpc_time>1189367718</rpc_time>
<timezone>28800</timezone>
<ncpus>8</ncpus>
<p_fpops>1887284643.824740</p_fpops>
<p_iops>3229719926.678640</p_iops>
<p_membw>1000000000.000000</p_membw>
<m_nbytes>1609011200.000000</m_nbytes>
<m_cache>1000000.000000</m_cache>
<m_swap>3597721600.000000</m_swap>
<d_total>10733223936.000000</d_total>
<d_free>4747952128.000000</d_free>
<n_bwup>609.364000</n_bwup>
<n_bwdown>0.000000</n_bwdown>
<avg_turnaround>13809.052334</avg_turnaround>
<credit_per_cpu_sec>0.004051</credit_per_cpu_sec>
<host_cpid>7c98db8e1d250deb1668b6008b53eabe</host_cpid>
</host>
<host>
<id>3769109</id>
<total_credit>730.097176</total_credit>
<expavg_credit>63.532120</expavg_credit>
<expavg_time>1189172640.702620</expavg_time>
<p_vendor>AuthenticAMD</p_vendor>
<p_model>Quad-Core AMD Opteron(tm) Processor 8347 [AMD64 Family 16 Model 2 Stepping 1]</p_model>
<os_name>Microsoft Windows XP Professional x64 Edition</os_name>
<os_version>, Service Pack 2, (05.02.3790.00)</os_version>
<create_time>1188992103</create_time>
<rpc_time>1189071715</rpc_time>
<timezone>-28800</timezone>
<ncpus>8</ncpus>
<p_fpops>1691278491.205370</p_fpops>
<p_iops>3617420121.779050</p_iops>
<p_membw>250000000.000000</p_membw>
<m_nbytes>4292640768.000000</m_nbytes>
<m_cache>250000.000000</m_cache>
<m_swap>4822474752.000000</m_swap>
<d_total>80014737408.000000</d_total>
<d_free>77291585536.000000</d_free>
<n_bwup>6074.430000</n_bwup>
<n_bwdown>0.000000</n_bwdown>
<avg_turnaround>26160.933424</avg_turnaround>
<credit_per_cpu_sec>0.003726</credit_per_cpu_sec>
<host_cpid>a4e0f7368478f121e59f9e5994f5f7fe</host_cpid>
</host>
<host>
<id>3773162</id>
<total_credit>1258.159284</total_credit>
<expavg_credit>104.570346</expavg_credit>
<expavg_time>1189341403.440610</expavg_time>
<p_vendor>AuthenticAMD</p_vendor>
<p_model>Quad-Core AMD Opteron(tm) Processor 8347 [x86 Family 16 Model 2 Stepping 1]</p_model>
<os_name>Microsoft Windows Server 2003</os_name>
<os_version>Enterprise Server Edition, Service Pack 1, (05.02.3790.00)</os_version>
<create_time>1189082449</create_time>
<rpc_time>1189367483</rpc_time>
<timezone>28800</timezone>
<ncpus>8</ncpus>
<p_fpops>1929784960.924450</p_fpops>
<p_iops>4248438434.327790</p_iops>
<p_membw>125000000.000000</p_membw>
<m_nbytes>4292640768.000000</m_nbytes>
<m_cache>125000.000000</m_cache>
<m_swap>6266802176.000000</m_swap>
<d_total>20973617152.000000</d_total>
<d_free>17219715072.000000</d_free>
<n_bwup>4425.880000</n_bwup>
<n_bwdown>0.000000</n_bwdown>
<avg_turnaround>6620.774901</avg_turnaround>
<credit_per_cpu_sec>0.003847</credit_per_cpu_sec>
<host_cpid>8a788a2ff5c503861bf10505f0ae1d8d</host_cpid>
</host>

There is also This system to draw statistics from.

Chart running. Now, if the people with nine-day caches would kindly stop extending them to ten days, at least until the server status page shows a healthy buffer again, we might get some meaningful data.

I must be missing something with the Barcelona, for the hype I think my boat anchor crunch's faster or almost anyway.

[edit]I guess that would make Barcelona as a quad, four boat anchors.[/eoe]

---

There is also This system to draw statistics from.

Chart running. Now, if the people with nine-day caches would kindly stop extending them to ten days, at least until the server status page shows a healthy buffer again, we might get some meaningful data.

Just human nature, anytime you come to the party so late, all of the best treats are already gone.
From what I see of Barcelona so far, their appitite is not going to have a great impact on available crunchable treats.

---

When we finally figure it all out, all the rules will change and we can start all over again.

I see that those times are crummy, considerably slower than those of my PD950s. And I wonder, even with Chicken's cruncher, whether those cores will even be as fast as mine.

---

I see that those times are crummy, considerably slower than those of my PD950s. But I also see he's (they're) using the default client(s).

The stock app (the Boinc client doesn't do the crunching) is not too shabby these days, so I would write it off to that.
Give the users some time to tweak their rigs and see if they start doing better.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

been waiting for barcelona so i could go get some serious crunching done,think i might get me com 64 out instead

yes and than on Linux64 bit...., with an optimized Client

i think there is going somethink

@who?

we know that you are working for Imhell, so be ....

Greetings from Germany NRW
Ulli

Well, I run 64bit Linux and optimized app on this one giving me about 15-20% advantage over win/stock-app.

Now do the calculation yourself...

Sorry, Ulli das sieht nicht gut aus.

I use both, Intel and AMD - and from my experience - Intel works better for Seti. On the other hand RieselSieve seems to favour AMD cpus...

thanks for the Info

hope that the Phenom comnes with 3 Ghz....

Greetings from Germany NRW
Ulli

I must be missing something with the Barcelona, for the hype I think my boat anchor crunch's faster or almost anyway.

[edit]I guess that would make Barcelona as a quad, four boat anchors.[/eoe]

or if its in a quad fx configuration 8 cores ;) (2 quads in 1 box :) )

---

Do you Good Search for Seti@Home? http://www.goodsearch.com/?charityid=888957
Or Good Shop? http://www.goodshop.com/?charityid=888957

I would like to see a app specifically compiled for the Barcelona's

---