Einstein statistics look up / Graph
12 Dec 2011 23:11:28 UTC
Topic 196101
(moderation:
Hi All
I am fairly new to all this crunching stuff, but seems to be fun.
At the moment I am wondering if there is a graph showing the teraflops / calculating / computer power which has been flown into the project. Something like a stock market chart showing overall computerpower over the last couple of years or so ?
I assume that there should be something like an increasing curve, with the constant newly and faster computers added ?
cheers
Einstein statistics look up / Graph
)
The second graph down in this link shows computational increase through time, right from the beginning of Einstein@Home. Enjoy!
http://boincstats.com/stats/project_graph.php?pr=einstein
ahhh thanks, goes over
)
ahhh thanks, goes over credits was looking for teraflops or similar ....well impressive :)
Credit is calculated based on
)
Credit is calculated based on the FLOPS that a computer is providing and the amount of time the computer provides at that rate, so the rate of credit accumulation is directly proportional to the processing speed of Einstein@Home. I am not sure what the calculation is which relates credit to computational power, so hopefully someone else can shed some light on it, but if you were to plot the credit increases as shown on the chart in a spreadsheet, find the equation of the curve, take the derivative of that and plot it, that would show you the change in computing power over time. Knowing that Einstein@Home is currently computing at around 570 TFLOPS, you could calculate the approximate computational power at any earlier time.
This is a roundabout way to do it, but it would give you what you're looking for :)
thanks for the explanation
)
thanks for the explanation
Hallo Savanti! Unfortunately
)
Hallo Savanti!
Unfortunately I can´t show graphs here now, but I have them.
Kellen is right. I you plot "Cumputing Capacity [TFLOPS]" from E@H Server Staus Page against "Recent Average Credit [Cobblestone]" from the project statistics - top field - from
here, you get by a least square fit the proportionality factor of 1,0198E-5 +/- 4,3E-9 [TFLOPS/Cobblestone]. That´s a very good fit (R^2=0,99989) from 629 data pairs since more than 1.7 years.
If you plot "Computing Capacity [TFLOPS]" aginst "Time [d]" and take a least square fit over the same time interval, you get a slope of 960,2 +/- 0,3 [GFLOPS/d]. From 1st July 2010 to 1st Jan 2012 the Crunching Capacity has increased from 250 tom 580 [TFLOPS], and I believe, we will pass the mark of 600[TFLOPS) within the next 2weeks. When will we reach the tremendous mark of 1 [PFLOPS]? Some time in spring 2013 ??? Let´s hope !!!
If you divide The "Computing Capacity [TFLOPS]" by the "Participants Active Last Week" from E@H Server Staus Page and plot this against time, you get from a linear fit over the same interval as above a slope of 17,6 +/- 0,6 [MFLOPS/(d*Participant)], about 1 [GFLOPS/(2month*Participant]].
I hope, this will answer some of your questions.
Happy crunching wishes
Martin