Einstein@Home has passed 800 Teraflops!
Submitted on 27 Nov 2012 10:23:15 UTC
Many thanks to the DC Russian Union for their ""Pulsar Rampage" computing challenge which helped to push Einstein@Home past the 800 Teraflop computing-power level for the first time in the history of the project!
Bruce Allen
Director, Einstein@Home
Language
Copyright © 2024 Einstein@Home. All rights reserved.
Comments
RE: Floating point speed
)
Translation, Seti@home is still down ...
For the past few days we have
)
For the past few days we have been granting enough credit to correspond to a Petaflop/s. It's just a matter of days for the gliding average to show this on the server status page.
Regarding ATLAS:
- The tape library has been moved to the basement in the office building; in fact it should be sitting right under my chair one floor down (we moved to a larger office in the corner of the building to accommodate Heinz-Bernd, too).
- The Linpac tests for the TOP500 list were made with 5368 compute cores. Atlas currently has 6720 cores in 1680 nodes
- Add 66 GPU nodes with 4 CPU cores and 4 NVIDIA Tesla cards (C1060 and C2050) each. Due to various software issues the GPUs are not used to full extent yet. In addition these nodes are the first to be shut down in case of problems with the cooling, which we had a lot this year.
- The extension ("Bauabschnitt #2") currently consists of empty racks. The cooling facility originally planned for this extension is currently cooling the older part, because the original cooling facility never fully worked. A "final" replacement is planned for 2Q next year (IIRC).
BM
Edit: by far the most increase in computing power comes from GPUs on the BRP search. It has more than doubled in the past ~2 months.
BM
RE: Regarding ATLAS:
)
With the updates do you know what is the actual capacity in TF?
Just as curiosity too, 1,000,000 E@H in credits is equivalent to about 5TF?
Congrats for the upcoming PenthaFlops mark break, realy amazing, and congrats, your servers realy have the capacity to feed ours hungry hosts. You guys in the lab make an excelent work too.
RE: RE: Floating point
)
It's back up now.
Our small team brake the mark
)
Our small team brake the mark of 4MM of daily credit production, about 20TFlops to help our E@H friends. Hope that makes the 1PFlops mark a little closer...
946.8 TFLOPS :-) Cheers,
)
946.8 TFLOPS :-)
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
First time we hit the border
)
First time we hit the border of 950 TFLOPS at 13:46 (UTC) (interpolated) today at a rate of 0,36[TFLOPS/h]. But tonight it will drop below this level, because of this peculiar daily sawtooth like behaviour of the total crunching power. About 1/2h before midnight (UTC) it drops by up to 5%. (I would like to know why this is.)
Let´s see when we will pass the 1 PFLOS! At the next weekend? Let´s hope!!!
Kind regards and happy crunching
Martin
RE: First time we hit the
)
We will pass it - by Tuesday
Our small team reaches almost
)
Our small team reaches almost another 4MM of daily credit production today, about 20TFlops more adeed to the E@H cause. The 1PFlops mark is a little closer... Hope our E@H friends apreciate that and our enemies hate! (if we have any of course, i don´t hope so, who knows?)
I made a graph yesterday, but
)
I made a graph yesterday, but the status page was out of date most of the time tonight:
http://crna.zapto.org/rrd/einstein/
I could do graphs of any other parameter as well if desired.
RE: RE: Of course there
)
What is so brilliant about BOINC is that even those who have only completed ONE workunit has added to the cause!