Syracuse University rises to the top!
Submitted on 28 Feb 2013 22:38:41 UTC
Thank you and congratulations to Syracuse University whose recent average credit value in excess of 3 million is one of the largest that we have seen in the project's history. This represents about 4% of Einstein@Home's total computing power. There must be a lot of Syracuse University campus computers crunching Einstein@Home! Without your help a few weeks ago, I don't think we wouldn't have made it past the 1 Petaflop mark.
Bruce Allen
Director, Einstein@Home
Language
Copyright © 2024 Einstein@Home. All rights reserved.
Comments
Syracuse University rises to the top!
)
For those who wonder, this is the list of the current top participants.
Cheers,
Oliver
Einstein@Home Project
I do hope someone has
)
I do hope someone has permission to run BOINC on them ?
Almost LOL !
dunx
LOL i'm guessing the IT staff
)
LOL i'm guessing the IT staff hs pushed down an invisible client to every unit connected to their WAN.
Grazie MIlle,sono contento di
)
Grazie MIlle,sono contento di essere utile per la causa...!!!onorato di ricevere questo ringraziamento da parte sua..!!un abbraccio dall'Italia...in fede Dario...!!
RE: For those who wonder,
)
Looks like I need one more GPU card to get in the Top 50
I hope they are doing this
)
I hope they are doing this with permission. I read about a guy who was criminally charged and held liable for all the hardware that failed due to having BOINC run on it. I used to work at a university and was never able to secure this permission from them.
RE: I hope they are doing
)
In that case, I suspect he was a victim of malicious prosecution along with everything else.
Yes, getting permission to run BOINC is important, and you must do that before installing BOINC on your employer's computers.
But running BOINC does not, of itself, cause hardware to fail. The two Q6600 hosts on my account have been running BOINC 24/7 for well over five years, with the minimum of attention - I'm using one of them to post this reply. At worst, BOINC can reveal the weaknesses of a poor procurement decision somewhere else within the organisation. The person in the story (I think I know the one you mean) was probably being grossly overcharged for the services of an external software house to come and run the 'uninstall' program on the affected computers.
I know the person who was
)
I know the person who was involved in that case - he was, and as far as I know still is, a member of the Anandtech Distributed Computing forums. His was indeed a case of gross over-prosecution and deception on the part of university officials over what was going on.
What it basically boiled down to was that he had permission from a previous IT Director and the new director decided he: a) didn't like the individual involved, and b) decided he want the DC project removed from the affected machines, but didn't trust the person to do it, so it was billed to a 3rd party. Later, the department head convinced officials to press criminal charges.
A classic case of making a mountain out of a mole-hill.
"Chance is irrelevant. We will succeed."
- Seven of Nine
For Syracuse I'd be inclined
)
For Syracuse I'd be inclined to put it down to having a world class physics department, including their gravitational wave research group, some LIGO luminaries, and several high powered computing clusters.
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
That's pretty cool.
)
That's pretty cool.
I have personally known prof
)
I have personally known prof Peter G.Bergmann of Syracuse University and edited the Italian edition of his book "The riddle of gravitation". But I don't know if he is still alive, that was many years ago.I think he would approve Einstein@home.
Tullio
cool story, bro :D No,
)
cool story, bro :D
No, seriously. Wish them even more crunching!
Awesome!!!
)
Awesome!!!
RE: I have personally known
)
Yeah he died back in 2002 and was born in the same city as I was.......he was there 43 years before me though.
I think he was about 87yrs old.
He was at Syracuse until about 1982
I hope I did not sound too
)
I hope I did not sound too negative with my other post. I put everything I can into running the BOINC projects and would do more if I could. For me, it is a way to make SOME contribution to science. an amateur physicist, astronomer, etc., and I am thankful to be able to make some small contribution to these and other fields. I just know how some administrators and directors can be on campus from personal experience. All the best to Syracuse!
Regards,
Mark W. Patton
Peter Saulson is at Syracuse
)
Peter Saulson is at Syracuse too. He wrote Fundamentals of Interferometric Gravitational Wave Detectors which I haven't yet stopped re-reading. I may actually understand this GW business one day! :-)
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Work together
)
Work together
I come from China,
)
I come from China, exhibitions
I'll see if I can dig up some
)
I'll see if I can dig up some pictures of Syracuse's green computing center, from my visit last summer. It's pretty cool with heat recycling and natural gas power cells.
I believe their IT department has written their own scripts to run E@H on all campus computers, something like 8-10,000 desktops.
Plus they have a cluster with 4,000 cores (I think) running Condor jobs crunching LIGO data.
I'll see if I can confirm or correct those numbers from my poor memory tomorrow.
Joe
RE: I'll see if I can dig
)
Are you on the east coast or the west Joe?
What made you visit the Syracuse computing center......other than it being fun for us mad scientists
I see you are #25 on the stats page here......not bad for just over 2 years here.
-Samson
Hi Samson, I'm on the left
)
Hi Samson,
I'm on the left coast, Los Angeles, but I was lucky enough to get a job in the LIGO collaboration. I've been to Syracuse a couple of times once for a get together of the Data Analysis Software Working Group and once to pick the brain of the world expert on Condor.
A good friend and fellow pilot teaches physics at SYR and was the one who introduced me to my boss.
My best Syracuse story (for mad scientists) is about their data center. The IT department built this big, fancy, super efficient data center. When it was complete they found that the racks and racks of machines they had didn't use enough power or generate enough heat to get them into the efficiency sweet spot. So to say they were accommodating to hosting the Gravitational Wave group's 2 clusters is an understatement.
I wish we all had that kind of problem.
Oh, I've also been to Milwaukee and somewhere have a picture of the rack that hosts E@H (which is still down while they rebuild a huge zfs storage appliance from backups). I'll try to dig that out even though it looks like every other rack mounted server setup.
Joe
A correction to my previous
)
A correction to my previous guess, the campus computers are about 8,000 cores not computers. And they have 2 clusters, still trying to verify core count.
WOW!
)
WOW!
RE: A correction to my
)
that's still a big amount...
The page I linked to earlier
)
The page I linked to earlier mentions
which would neatly explain any E@H usage, as our executables can usually be run within such arrangements. Our WU's have been used for 'burn in' or testing of cluster components in other similar situations.
Cheers, Mike.
( edit ) On this page is a neat 28 minute video about their Green Data Center which houses much of their IT stuff. As you'd expect cooling and power are the key issues. Calculating machines are still thermodynamic devices and need due care .... lessons for the home user too. I like that they produce DC power onsite. Clever. :-)
( edit ) More info here. Basically the gravitational wave community are revving up their tools in expectation of Advanced LIGO data streams.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
An Italian firm, Eurotech, of
)
An Italian firm, Eurotech, of Amaro, Friuli, has built a computer rack which, according to theregister.co.uk, gets up to 3150 MFlops/watt with liquid cooling.The rack, called Eurora, is being tested at the CINECA computing center, the most powerful in Italy.If verified , it could be used to build a "green" supercomputer.
Tullio
RE: An Italian firm,
)
Impressive. Here is the news release from NVIDIA on this subject: http://nvidianews.nvidia.com/Releases/NVIDIA-GPU-Accelerated-Supercomputer-Sets-World-Record-for-Energy-Efficiency-910.aspx
Cheers
HB
To reduce confusion on the
)
To reduce confusion on the phrase 'hot water cooling' used for that setup : 'hot' is above 50C and thus is apparently (a) less then the chip/die temperature and (b) greater than environmental temperature. That way you get passive cooling ( heat flowing from a higher temperature object to a lower temperature object ) both at the chip face and at whatever is outlet of your system. So while the water is not necessarily re-circulated, the temperature differential on the outlet is available to drive some other process. You could heat some adjacent building for instance. Hence you can get several bites of the cherry before entropy says no. I think this implies that greater cooling - for the electronics - would be achieved by greater flow, so I'd expect that implies a decent supply of water.
[ This has some similarity with 'water hammer pumps' that take the kinetic energy of say, nine-tenths of some mass of water falling down to pump up the remaining one-tenth. It's an old technology but a goodie. ]
Mind you that leaves the question of where you got your 50C water from ... so I do hope the energy efficiency calculation accounts for that, as opposed to merely externalizing.
Cheers, Mike.
( edit ) Here's more on that. So you do recirculate. That's rather like my wee Corsair H-80 CPU cooler then ...
( edit ) I guess there is a natural tension that emerges when you mention 'water' adjacent to 'electricity' :-)
It says "The water in the circuit can be as hot as 50 °C .. " so that suggests that it is expected to work effectively even if lower. That sounds more like it. Thus it's the tremendous heat capacity of water and the relative ease of shifting it from the hotter spot to the cooler spot ( c/w air as a working fluid ) that makes the design. Doesn't sound especially unique ....
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Hallo Freunde, Ich war in
)
Hallo Freunde,
Ich war in Urlaub. Deshalb Keine Arbeit
Viele Grüße Werner
RE: I have personally known
)
Peter Bergmann was one of Albert Einstein's last research assistants, and worked with Einstein in the 1940s. I think Bergmann died in 2002, and was awarded the American Physical Society's first Einstein Prize posthumously.
Director, Einstein@Home
Thanks, dr.Allen.I think I
)
Thanks, dr.Allen.I think I met him in Florence in 1969 and published an article of his in the 1970 Mondadori Yearbook of Science and Technology. The article covered also the experiments of prof. Joseph Weber. I later learned that Edoardo Amaldi had visited Weber in the USA and this gave rise to the Italian line of research on gravitational waves,which culminated in VIRGO.
Tullio
I had a chance to talk to one
)
I had a chance to talk to one of the people from the Physics Dept. involved in setting up thier cluster in the Syracuse data center.
First of all I was wrong (surprise, surprise) about the power generation. It is done with natural gas powered turbines not fuel cells.
The cooling device is an abosrbtion cooler which uses Lithium Bromide to turn waste heat into chilled water. The actual device is http://www.thermaxindia.com/Absorption-Cooling/Products/Vapour-Absorption-Machines/Steam-Driven-Vapour-Absorption-Machine.aspx
The way he explained it to me in the winter outside air is used to cool the recirculating water and the heat from the turbines is used to warm the nearby buildings.
In the summer the heat from the turbines goes into this Absorbtion Cooler (Wikipedia has a good description of the technology) which chills the water. It turns out to be much less efficient than normal freon or A5 (I think) air conditioners. But when the heat is free the equation changes.
The chilled water is then circulated to heat exchangers on each rack and the fans in the computer racks push the hot air over the heat exchangers. So there is no air conditioning running under normal circumstances in the building.
Joe
RE: I later learned that
)
I think this is correct. The Italian line of research began with the work of the Prof. Guido Pizzella, at Università degli Studi di Roma La Sapienza. Indeed, Pizzella was a protege of Amaldi.
You might be interested to know that there is a direct connection to Einstein@Home. Maria Alessandra Papa, who is one of the central people behind the Einstein@Home search for continuous gravitational waves, comes from this Italian research group (PIzzella was one of Marialessandra's two Phd thesis advisors).
Bruce
Director, Einstein@Home
Allow me a small correction.
)
Allow me a small correction. Guido Pizzella. As a physics and astronomy editor at Mondadori I have published articles by Bergmann, Remo Ruffini and Bruno Bertotti on experimental general relativity. One of my advisers at Mondadori was prof. Carlo Castagnoli of Torino University who was the leader of the neutrino detectors group in the Mont Blanc tunnell during the observations of supernova 1987A
Tullio
RE: Allow me a small
)
Sorry, thank you, I have fixed this!
I'm glad to have you involved in this project as well!
Bruce
Director, Einstein@Home
Thanks. I have always been
)
Thanks. I have always been interested in General Relativity but when I graduated in 1967 in Theoretical Physics in Trieste it was not followed much, and I had to look for a way to make a living. I found it in Milano at Mondadori and came to know a good number of astronomers and physicists working in that field, especially Remo Ruffini who started a cooperation with my friend Gallieno Denardo. But since I left Mondadori in 1979 I lost touch with them. No Internet then.
Tullio
Well I've been writing user
)
Well I've been writing user documentation for a while so all my "one of these days" projects are getting checked off.
Here are a few pictures from Syracuse and Milwaukee that you may find interesting. I apologize for the amateur quality they were taken with my phone and the lens was not as clean as it should have been.
A room full of gatural gas turbine generators
------------------
The controls and read out of one of the turbines. I'm always amazed at how fast these things spin.
---------------------
The power output monitor. I'm not sure if it's looking at one or all.
------------------------------
A small part of the plumbing with blue cold water, orange hot, and I think the pale blue is the cold water return from the server racks.
---------------------------
This is one of the monitor/control panels for an absorption cooler.
--------------------------
A random picture of the Syracuse University campus
------------------------------
This one is at U of Wisconsin at Milwaukee. One of the computers in the middle rack is einstein.phys.uwm.edu. The silver covered device is a raid array holding one copy of our data and results.
Joe
RE: Well I've been writing
)
THANK YOU for the pictures!!! It is always cool to see what can be done!!
Hey as an aside next time you are in Syracuse try the place "Funk-n-Waffles" it was featured on the show Diners, Drive-ins and Dives and they serve waffles as the base for lots of things, pretty cool!
http://find.mapmuse.com/details/diners-guy/999153552/funk-n-waffles