The Large Hadron Collider project, under construction at CERN in Geneva Switzerland, is the largest project in High Energy Physics to date, if not one of the largest scientific projects undertaken by humankind. The two main high energy detectors are ATLAS and CMS. Here are a few interesting facts about the numerology:
- The well-known “circular” collider is actually not a circle. It consists of 8 straight sections and of 8 arcs.
- The rate of proton-proton interactions at design luminosity is 109 events per second.
- The weight of the CMS experiment is roughly 13000 tons (which is 30% more than the Tour Eiffel) .
- The ATLAS experiment is reputed to be able to float in water.
- The amount of cables used in ATLAS is roughly 3000 km.
- The data volume recorded at the front-end in CMS is 1 TeraByte/second which corresponds to 10,000 sets of the Encyclopedia Britannica. Per second.
- The data recorded during the 10-20 years of the LHC lifespan will be equivalent to all the words spoken by humankind since its appearance on earth.
- The kinetic energy the beam is equivalent to 1 small aircraft carrier weighing 104 tons traveling at 20 miles/ hour.
- The machine temperature is 1.9 Kelvin (which will be the largest cryogenic system in the world).
- The total cost of machine + experiments is roughly 5000 MCHF (or 8 billion US dollars, using US accounting techniques).
- The total number of physicists involved is roughly 5000, a large fraction of which are from the US.
September 27th, 2006 at 1:22 am
“The kinetic energy the beam is equivalent to 1 small aircraft carrier weighing 104 tons traveling at 20 miles/ hour.”
An aircraft carrier is closer to 50,000 tons. I think you forgot to include the first order corrections to the approximation.
September 27th, 2006 at 1:47 am
Since your doing LHC Factoids I thought those with an lay person interest might like, All that you can know about Cern in seven questions?
September 27th, 2006 at 3:27 am
I was slightly confused by your use of the word “factoid”. I’m not a native speaker, but I’m sure that it’s usually used to mean something tht is not true but maybe believed to be. (Looking it up I find this discussion of the word.) The first point in the list also has this form, but then comes things that look like plain facts and I have to rethink what the post is about. Just a small note, so you know that this might not be immediately clear.
September 27th, 2006 at 5:00 am
I think you’ve got a bit of a superscript problem – 10^9 interactions/sec, 10^4 ton aircraft carrier!
September 27th, 2006 at 5:26 am
“The total cost of machine + experiments is roughly 5000 MCHF or 8 billion US dollars.”
5000 MCHF is just over 4 billion USD (not 8 billion). Unless you’re making a commentary on the different accounting systems used by the US and Europe….
September 27th, 2006 at 10:05 am
What I found most amazing when I visited CMS and ATLAS was that the colors of the detectors actually are the same as in the simulation plots. I mean, all that bright green/red/blue, like made from a children’s construction kit or so
September 27th, 2006 at 11:02 am
Scott errors in metric and displacement measures or other naval stats may prove to be purely or wholly or ‘intentional’
-
I note in english we call it the Eiffel Tower and the cycling the Tour de France,
but hey when the data recorded during the 10 to 20 years (+/-1 or 2) of the LHC lifespan will be equivalent to all the words spoken by humankind since they started talking, one needs to bear in mind much has been said in many languages (and a lot in error). That’s an awful lot of hot air!
Maybe they should call LHC the prolific swiss Shakespeare. I guess the data just from the LHC should keep several generations of scientists and students busy for the next 500 years. Give or take one or two (years).
September 27th, 2006 at 11:16 am
“The data volume recorded at the front-end in CMS is 1 TeraByte/second which corresponds to 10,000 sets of the Encyclopedia Britannica. Per second.”
How about “The data volume recorded at the front-end in CMS is 1 TeraByte/second which corresponds to 1 terabyte/second”?
For christs sake. This is 2005, and scientifically literate people are reading this. If someone is so computer illiterate that they can’t understand what a TB/s means, how is talking about Encypledia Britannica sets going to add to the conversation?
September 27th, 2006 at 11:28 am
Maynard, have you thought about switching to decaf?
I for one have a much better intuitive grasp of the information content in the Encylopedia Britannica than in one terabyte.
September 27th, 2006 at 11:46 am
Come on, Sean, you can not be serious. You have a computer, right? You have a hard drive in it, right, storing what, 300GB? So a TB is 3.5 the size of that hard drive. You seriously want me to believe that you can visualize 10,000 copies of the Enc Britannica, but you can’t visualize 3.5x the storage of your hard drive?
September 27th, 2006 at 11:55 am
Scott: these factoids were written by folks at CERN. Perhaps Swiss aircraft carriers are smaller than American ones.
Anna: Here’s the entry from Merriam-Webster:
Main Entry: fac·toid
Pronunciation: ‘fak-”toid
Function: noun
1 : an invented fact believed to be true because of its appearance in print
2 : a briefly stated and usually trivial fact
Clearly, I am employing the second usage here.
Mike, Yep, I meant in US accounting – I’ve fixed the text to make it clear.
Maynard, get a grip on it! If you want complain about something important, try the Iraq war.
September 27th, 2006 at 12:06 pm
[...] Read More at Cosmic Vaviance [...]
September 27th, 2006 at 12:22 pm
Look, the reason for complaining about the Enc Brit comparison is that it is one more form of innumeracy. You would, and rightly so, criticize a journalist who wrote something innumerate.
A useful comparison of this sort compares like with like in some way.
I could say “well a TB/sec is like going 100m/s if a byte were an angstrom”, but that’s a completely stupid comparison. I could say “well a TB/sec is like eating one hamburger/second if a hamburger were a TB” which is, in point of fact, exactly as stupid as the Enc Brit version. A hamburger has a vast amount of “information” in it, if read properly. The Enc Brit carries a certain image of size and heft, but that has nothing to do with what humans call the information in it, it has to do with the paper and ink and their irrelevant specific molecular configurations. The Enc Brit analogy is every bit as misleading as, say, a graph that draws 3D objects not bars, so that a value twice as large as another gives an impression of being eight times as large — it specifically evokes an image of size and weight that is not relevant to the issue.
A valid sort of analogy would focus on the real issue here which is the data rate. A current PC depending on cost and configuration, can transfer data to/from the hard drive at between 10 and 100MB/s. Thus we are talking about data transfer rates that are between 10,000 and 100,000 times the rate aot which your PC can store data to its hard drive. This is a comparison that is salient (like compared with like) and therefore useful.
As for the Iraq war; really, Joanne? You feel that it would be appropriate of me to comment, in a thread about the LHC, on the Iraq war? If you want my opinions on the Iraq war, feel free to look up my name in google or technorati or some such (hint, I was against it before it started); but I think most people would agree that blog comments are best served by staying on topic.
Look, I know I am abrasive. But, honestly, can you read my point about the correct way to analogize this data rate and tell me, to me face, that I am wrong?
September 27th, 2006 at 12:31 pm
You are wrong.
September 27th, 2006 at 2:01 pm
Maynard, Sean is correct – you are wrong. If one were to type, say into Word, the knowledge contained within the pages of the Encyclopedia Britannica, and repeat this 10,000 times, the file size would be 1 Terabyte. That’s all there is to this analogy – it’s a perfectly valid comparison that most people can understand. The folks at CERN checked it out.
As to the Iraq war, obviously I do not expect such comments on an LHC thread, and that was not my point.
September 27th, 2006 at 2:39 pm
The Encyclopedia Brittanica comparison is cliched, not wrong. It’s used every time somebody tries to convey the sense of a big data rate or volume. Perhaps it would be more interesting/useful to compare the LHC data rate to the number of times your genome could be sequenced per second, or the number of Google searches you could do per second, or the number of simultaneous phone calls the NSA could listen in on at one time?
September 27th, 2006 at 2:52 pm
The displacement of the first commissioned aircraft carrier, the Hosho, was 10,500 tons, fully loaded so 10^4 is about right. But as far as the ATLAS experiment floating; I wonder if the authors of that factoid think that the Titanic floats.
September 27th, 2006 at 3:09 pm
I’m coming down against Maynard. I have a set of the Britannica at home, so if I imagine all the data of 10,000 such sets piped across each second, I say “wow, that’s a lot of data”. The problem is that one hard drive looks pretty much like another. Today’s 250 GB drives are about the same size as 20 GB drives a couple years ago, so I don’t immediately connect the storage capacity of hard drives to a large number the way I can with something that takes up real space. Call it simple-minded, but analogies to readily available spatial things are really effective, as long as the numbers are accurate.
September 27th, 2006 at 3:46 pm
Maynard a dit:
>This is 2005, and scientifically literate people are reading this.
>…But, honestly, can you read my point about the correct way
> to analogize this data rate and tell me, to me face, that I am wrong?
Well, you’re definitely wrong about what year it is.
September 27th, 2006 at 3:57 pm
I also like the Encypledia Britannica analog. Talking abouyt the amount of information in a hard disk is, at least, difficult: how much is redundant information? how do you account for the fact that for computer data you’ ve got different compressing algorithms?
Which leads me to think in really how much of the info recorded will just be garbage. I mean, is just a collider, is it really possible that the info collected from it is more than the actual knowledge (or at least the spoken) of humanity? doubt it.
Good article, though. But maybe the sources of the data have a too big ego..
September 27th, 2006 at 4:13 pm
Actually I will say that I have a hard time with visualizing how many books 10,000 Encyclopedia Britannicas that is. It would be helpful if someone could tell me how many sets of Encylopedia Britannicas could fit between here and the moon. Or alternatively how many sets of 10,000 could fit between here and the moon? Anyone?
September 27th, 2006 at 4:14 pm
Corrrr Blimey! Proof if ever was needed, that someone is always watching and any and all errors can be (and are) promptly challenged by peer review in the scientific community. Almost toooo scared to make a comment.
-
Maynard getting the year wrong is definitely a no no, if you are going to challenge the accuracy of a report.
JoAnne 10 (to the power of 4) tons is still a very small frigate, but do the swiss even have a navy??? lol!
Carl the Titanic did float ‘originally’ it just hit an iceberg, and twin hulls hadn’t been made compulsory yet. I hope CERN haven’t scrimped and saved on the thickness of their hull. Not that there are many icebergs to watch out for nowdays, floating around Geneva.
-
And finally cliche or no cliche, if it can record ten thousand encyclopedia britannica’s every second, how many scientists does it take to read the 60 Terabytes or 600,000 encyclopedia britannicas recorded in 1 minute
September 27th, 2006 at 4:16 pm
When the experiments at the LHC have run their course, how many words will be added to the Encyclopedia Brittanica to reflect the new knowledge?
September 27th, 2006 at 4:43 pm
And …. and ….
Apropros of nothing except monkey banging on typewriters …
…. at 10,000 Encyclopedia Brittanicas per second how long until the LHC writes the actual Encyclopedia Brittanica. Anyone?
September 27th, 2006 at 5:04 pm
To me, the most interesting fact is the tiny fraction of events that will actually be recorded. It’s 10^9 events per second, and 10^3 GB per second, but as I recall it’s something like 10 GB to record a single event, so only 100 of those billion events per second are being kept for analysis. The rest are quickly perused by the detector and tossed out for being not worth our while. At some point we should post something more detailed about this whole concept of a “trigger,” as I suspect that most non-high-energy-physicists would be surprised to hear that the LHC throws away the vast majority of its data.
September 27th, 2006 at 5:30 pm
Sean (#25), You’ve hit upon one of the most interesting and challenging and potentially frightening aspects of the LHC. We had a great lecture on the triggers at the SLAC Summer Institute and I’ve started a draft of a post on the subject. This gives me the push I needed to finish it off!
September 27th, 2006 at 7:49 pm
MCHF? Mega-Confoederatio-Helvetica-Francs?
September 27th, 2006 at 9:01 pm
“For Christ’s sake. This is 2005, and scientifically literate people are reading this. If someone is so computer illiterate that they can’t understand what a TB/s means, how is talking about Encypledia Britannica sets going to add to the conversation?”
No it makes perfect sense to make this reference. A terabyte, even for the smartest people, is hard to conceptualize. It is like when someone says “American is 1trillion dollars in debt” that number is so huge its worthless and wasted on the brain.
But, people have seen the encyclopedia; they know its huge, and 10,000 is a number we can almost visualize (though a secondary size comparison might help — how many libraries of congress would that fill?)
Also, it attaches a sense of emotion to it. I have never encoded data, but I have written a paper. I have spent hours and hours writing 20page papers. And that is next to nothing compared to one letter in the encyclopedia. Now the whole thing? 10000 times? A SECOND?!?!?
Emotive, interesting, comprehendible.
Oh and, its 2006
Oh, and you spelled “encypledia” wrong. Its 2006 man, where is your spell check?
September 27th, 2006 at 9:17 pm
I’m sorry, I’m a software engineer. I am not familar with this Britannica thing. How many copies of wikipedia (all languages) fit into one Terabyte?
September 27th, 2006 at 9:40 pm
JoAnne (#15):”If one were to type, say into Word, the knowledge contained within the pages of the Encyclopedia Britannica, and repeat this 10,000 times, the file size would be 1 Terabyte.”
I’m not sure this is such a good idea. Word is a bit of a bloated format. How many bytes would 10,000 Encyclopedia Britannica’s be in ASCII??
Also, in response to spatulated (#28), according to Wikipedia 1 Library of Congress = 67,000 Britannica’s. I guess the data volume recorded at the front-end of CMS corresponds to about 9 Library of Congress’s per minute.
September 27th, 2006 at 9:57 pm
“9 Library of Congress’s per minute” wins. Although I would round up to 10.
September 27th, 2006 at 11:19 pm
The correct comparison for data storage would be in units of LoCs, 1 TB/s ~ 0.05 LoC/s. For the European crowd, LoC = Library of Congress and the volume of print data it holds. From wikipedia: Occasionally, this figure has been referred to as a data transfer rate, LoC/s—Libraries of Congress per second — defined as 20 terabytes of data transferred per second.
And what ever you do.. don’t lick the beam pipe
September 27th, 2006 at 11:33 pm
All of these analogies with Encylopedia Britannica are just so 1990s… If a DVD can hold a two hour movie in, say, 8 gigabytes, then one terabyte/second is about 125 full-length features per second. Now THAT’S a lot of data.
September 28th, 2006 at 12:18 am
You are so missing the point, with all this discussion about Encyclopedia Brittanica and terabytes… Last week I saw the ATLAS and CMS detectors “up close and personal” for the first time, and it suddenly sunk in, what amazing discovery potential is opening up in front of us in a year’s (or two) time. And also, although I’ve been doing particle physics for all my adult life, I couldn’t help but being awed as if it were the first time by the human ingenuity that has devised and built these extraordinary machines in order to study the tiniest of particles.
Just a year away from the LHC starting date and all that I hope we will learn from it, I am so excited and feel lucky to live and be doing particle physics at this pont in time!
September 28th, 2006 at 12:35 am
Sean,
I believe you’re a few orders of magnitude off on the event size. If I recall correctly, ATLAS events are about one megabyte each, corresponding to order of 100 MB of data being stored per second. This gives me 100*10^7 MB = 1000 terabytes a year. (Experimentalists usually call a year 10^7 seconds, rather than 3.15*10^7, because we’ll be lucky to collect data that fraction of the time.) That’s a pretty big strain on available computing power as it is—keep in mind that we don’t just need to store all that data, but make it accessible to all the collaboration members who want to analyze parts of it.
September 28th, 2006 at 4:46 am
Hi Marty, I like that
125 DVDs or full feature length movies a second.
7500 movies recorded in 1 minute. Every minute.
-
I guess scientists who review the data (movies) could be called film critics.
That’s a lot of movie (data) watching.
-
Mind you 10 libraries of Congress every minute that is definitely gonna take some reading. So these triggers will be like librarians or clerks in the patents office.
Perhaps the triggers should be called little Einsteins
-
September 28th, 2006 at 6:13 am
[...] På den fantastiske blog Cosmic Variance har jeg fundet dette indlæg, som fortæller lidt facts om den enorme partikelaccelerator, Large Hadron Collider, som er under opførelse på (eller retttere under) CERN i Schweiz. [...]
September 28th, 2006 at 12:47 pm
Anna: you go girl! I couldn’t have said it better. And, yes, actually seeing the detectors is awe inspiring. I saw CMS last summer and hope to get to CERN sometime before next summer to see ATLAS and CMS underground, before everything is locked up.
September 28th, 2006 at 1:24 pm
While most of us will remain at a distance from this research and development, we can learn much by the factoids and links too, what is accessible and produced for the public?
Beyond Einstein, and from Seti it was learnt? Why not use “other” computers?
You take the chance on blogging, to see the benefits later on?:) A new student , a lay person, who now can’t get enough. More then, just the “pizza guy” and or the “toppings” for sure. And we have this developing theory from what has already been produced in experimental research?
How appropriate.
http://pancake.uchicago.edu/%7Ecarroll/images/cms1.jpg
Sean Carroll:
September 28th, 2006 at 3:43 pm
JoAnne, you said “… We had a great lecture on the [ LHC ] triggers at the SLAC Summer Institute and I’ve started a draft of a post on the subject. …”.
The pdf by Paris Sphicas at
http://www-conf.slac.stanford.edu/ssi/2006/lec_notes/Sphicas072606.pdf
is very interesting, and I am sure that you heard much more at the actual talks and discussion, so I am looking forward to your post on triggering.
Will you also discuss cuts, another important stage of filtering what will actually be analyzed at LHC?
Tony Smith
http://www.valdostamuseum.org/hamsmith/
September 28th, 2006 at 9:20 pm
[...] 12 – LHC Factoids “The well-known “circular” collider is actually not a circle. It consists of 8 straight sections and of 8 arcs.” Gen up on the Large Hadron Collider at Cosmic Variance! (tags: particle physics science trivia data misconceptions myths facts Collider Hadron Large LHC) [...]
September 29th, 2006 at 1:18 am
I know most of you are sick of those Encyclopedia Brittannica comparisons by now, but a while back I made a little webpage on quantities of information, and I’m not going to let it go to waste:
# A gigabyte is about a billion bytes (actually 2^30 of them).
* 1 gigabyte: the human genome, or a pickup truck full of books.
* 20 gigabytes: a good collection of the works of Beethoven.
* 100 gigabytes: a library floor of academic journals.
# A terabyte is about a trillion bytes (actually 2^40 of them).
* 2 terabytes: an academic research library.
* 6 terabytes: all academic journals printed in 2002.
* 10 terabytes: the print collections of the U.S. Library of Congress.
* 40 terabytes: all books printed in 2002.
* 50 terabytes: all mass market periodicals printed in 2002.
* 60 terabytes: all audio CDs released in 2002.
* 80 terabytes: capacity of all floppy discs produced in 2002.
* 140 terabytes: all newspapers printed in 2002.
* 170 terabytes: the searchable portion of the World-Wide Web in 2002.
* 250 terabytes: capacity of all zip drives produced in 2002.
# A petabyte is about 10^15 bytes (actually 2^50 of them).
* 1.5 petabytes: all office documents generated in 2002.
* 2 petabytes: all U.S academic research libraries.
* 6 petabytes: all cinema release films in 2002.
* 20 petabytes: all X-ray photographs taken in 2002.
* 90 petabytes: the “Deep Web” in 2002.
* 130 petabytes: capacity of all audio tapes produced in 2002.
* 400 petabytes: all photographs taken in 2002.
* 440 petabytes: all emails sent in 2002.
# An exabyte is about 10^18 bytes (actually 2^60 of them).
* 1.3 exabytes: capacity of all videotapes produced in 2002.
* 2 exabytes: capacity of all hard disks produced in 2002.
* 5 exabytes: all the words ever spoken by human beings.
* 6 exabytes: information in the genomes of all the people in the world.
You may think you don’t need to look at my webpage now, but you’re wrong! You have to look at it to see how much information it takes to completely describe the quantum state of a typical raindrop.
September 29th, 2006 at 8:56 am
I mean for sure we tend to see this “complexity of the information” and let’s not loose sight of the “powers of ten?”
For the lay person like myself this compartive relation of information dropped down to “scale” is important. As it puts into perspective the “reality” around us.
The value of the Qubit, and the lesson we may learn on how computerization may now meet new ways to decipher LIGO INformation?
So I find it difficult to image the relation in the complexity of the “rain drop” too a string? Yet, LHC serves it’s purpose from the bottom up?:)
September 29th, 2006 at 12:13 pm
Just to add a couple of factoids. The CMS magnet operates at 4 Tesla. The energy stored in the magnetic field is about 2.5GJ. If this energy were converted to potential energy, it could lift the whole CMS experiment about 20 m. Or, it could lift 28 tons to the top of Mt Everest…
And it is true that ATLAS can float. However, you would have to wrap it in Saran wrap (a lot of Saran wrap!) first…
The trigger for the LHC is a fascinating and, to some extent, scary topic. In my previous experiment at LEP, we accumulated about 4 million hadronic Z decays (among other events). That was essentially all of the produced Z bosons. We only missed a couple thousand. And we even have a pretty good idea why we didn’t trigger on those events. In other words, our acceptance was about 100%. At the LHC, the number of potentially interesting events produced will simply be too high: on the order of 10 to 100 thousand events per second. We only have the capability to record data at 100-300 Hz. Thus we have to throw away the least interesting events, while ensuring that we record the really exciting ones like the Higgs events (should they exist). A daunting task!
October 4th, 2006 at 5:03 pm
[...] The Large Hadron Collider accelerates protons to an energy of 7000 GeV, which is pretty impressive. (A GeV is a billion electron volts; the energy in a single proton at rest, using E=mc2, is about 1 GeV.) But it requires a 27-kilometer ring, and the cost is measured in billions of dollars. The next planned accelerator is the International Linear Collider (ILC), which will be similarly grand in size and cost. People have worried, not without reason, that the end is in sight for experimental particle physics at the energy frontier, as it becomes prohibitively expensive to build new machines. [...]