37.1 TFlops
log in

Advanced search

Message boards : News : 37.1 TFlops

Author Message
Profile Kevin
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 27 Jul 12
Posts: 507
Credit: 14,344,251
RAC: 1,727
Message 1643 - Posted: 19 Sep 2013, 3:21:27 UTC
Last modified: 19 Sep 2013, 3:30:36 UTC

In more serious news, as of this morning (Perth Time), we are processing at 37.1TFlops a new record.

Thanks, Merci, Danke, Спасибо за поддержку, Arrr that be grand.
____________
Regards
Kevin
-----
International Centre for Radio Astronomy Research

Alez
Send message
Joined: 12 Aug 12
Posts: 20
Credit: 21,130,689
RAC: 0
Message 1651 - Posted: 22 Sep 2013, 12:08:12 UTC

39.229 TFlops today.
Challenge on at Boincstats http://boincstats.com/en/stats/challenge/team/chat/417
Will 40 TFlops be broken ?

USTL-FIL (Lille Fr)
Send message
Joined: 4 Sep 13
Posts: 3
Credit: 186,345,660
RAC: 5,792
Message 1654 - Posted: 23 Sep 2013, 13:36:22 UTC - in response to Message 1651.

39.229 TFlops today.
Challenge on at Boincstats http://boincstats.com/en/stats/challenge/team/chat/417
Will 40 TFlops be broken ?


Done!:
GigaFLOPs actuels 40,655

Profile Steve Hawker*
Send message
Joined: 31 Oct 12
Posts: 6
Credit: 1,630,267
RAC: 0
Message 1655 - Posted: 24 Sep 2013, 1:14:28 UTC - in response to Message 1654.

On the Skynet site it says 44.84 TFLOPs

Profile alexbeckley
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Jan 13
Posts: 345
Credit: 5,696,421
RAC: 0
Message 1656 - Posted: 24 Sep 2013, 1:30:30 UTC - in response to Message 1655.

On the Skynet site it says 44.84 TFLOPs

That includes the processing from the SourceFinder project as well as POGS.

Alex

Cameron
Send message
Joined: 20 Oct 12
Posts: 19
Credit: 508,089
RAC: 6
Message 1657 - Posted: 24 Sep 2013, 1:52:21 UTC

Obviously you're happy with whatever average processing power is donated to the project.

Since it was mentioned that prior to L'Alliance Francophone abut 7% of the data set had so far been processed (7K of 100K) and at around 28 Tflops there was a further 2-3 years needed to complete the data set.

I was curious about what may be planned as we approach the end of the data set although we may still be a little while before that happens.

-- Cameron

Profile David Thilker
Volunteer moderator
Project tester
Project scientist
Send message
Joined: 29 Jul 12
Posts: 11
Credit: 9,820,491
RAC: 2,164
Message 1660 - Posted: 27 Sep 2013, 3:30:48 UTC - in response to Message 1657.

In answer to your question, Cameron: As you probably guessed, the exact number of galaxies to be processed depends on the computing power donated. The figure of 100K was an initial estimate, which we hope will end up being on the low end as the project matures.

Also, complicating the matter somewhat is the fact that we can process some galaxies in more than one form: at high angular resolution using only the optical data from PS1 or SDSS, and at lower resolution including the observations in the UV from GALEX and in the IR from WISE.

In the long term, after the basic survey is done, I look forward to applying the same pixelSED techniques to the high redshift universe being probed by sensitive Hubble Space Telescope imaging campaigns, such as the Hubble Ultra Deep Field and complementary projects like CANDELS (http://candels.ucolick.org).

With your help (and that of everyone contributing), all this becomes possible. Thank you!

mpolymorphe
Send message
Joined: 4 Oct 12
Posts: 1
Credit: 328,208
RAC: 0
Message 1682 - Posted: 7 Oct 2013, 21:37:49 UTC - in response to Message 1660.

No plan to use GPU?

Profile Kevin
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 27 Jul 12
Posts: 507
Credit: 14,344,251
RAC: 1,727
Message 1689 - Posted: 9 Oct 2013, 1:10:59 UTC - in response to Message 1682.

I'm in the throes of porting the F77 to C so I can move the code to a GPU.
____________
Regards
Kevin
-----
International Centre for Radio Astronomy Research

JugNut
Send message
Joined: 10 Aug 12
Posts: 22
Credit: 62,403,379
RAC: 8,931
Message 1692 - Posted: 9 Oct 2013, 13:07:44 UTC - in response to Message 1689.

Well done Kevin. Can't wait to get a crack at some of those..

Any time line as yet? Perhaps?:)

adrianxw
Send message
Joined: 16 Aug 12
Posts: 26
Credit: 7,019,455
RAC: 10,362
Message 1696 - Posted: 16 Oct 2013, 9:51:26 UTC

The f2c fortran to c convertor may save you some time, http://www.thefreecountry.com/compilers/fortran.shtml. Not used it myself, could be more trouble than it is worth, these things often are.

I'd offer to help you myself, (20 years F77, engineering apps on DEC and SEL mini's, 10 years C/C++ on Windows/Linux/embedded systems), but have not programmed since 2008.

Profile Kevin
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 27 Jul 12
Posts: 507
Credit: 14,344,251
RAC: 1,727
Message 1697 - Posted: 16 Oct 2013, 23:18:18 UTC - in response to Message 1696.

I've looked at the f2c tool - unfortunately the output isn't particularly readable
____________
Regards
Kevin
-----
International Centre for Radio Astronomy Research

adrianxw
Send message
Joined: 16 Aug 12
Posts: 26
Credit: 7,019,455
RAC: 10,362
Message 1698 - Posted: 17 Oct 2013, 11:30:47 UTC

Mmm, I suspected as much.

Ananas
Send message
Joined: 13 Oct 12
Posts: 36
Credit: 2,500,031
RAC: 0
Message 1699 - Posted: 18 Oct 2013, 7:53:31 UTC - in response to Message 1696.
Last modified: 18 Oct 2013, 8:07:42 UTC

... I'd offer to help you myself, (20 years F77, engineering apps on DEC and SEL mini's, 10 years C/C++ on Windows/Linux/embedded systems), but have not programmed since 2008.

OT:
SEL is Standard Electrik Lorenz? I didn't even know that they made computers. DEC I know of course, e.g. Rainbow 100 with cp/m 86 and this funny double floppy drive - one slot normal orientation, one upside down.

As C is a way of thinking, a concept rather than just a language, I highly doubt that you ever will forget how to do it :-) You might have forgotten some library function syntax but that can easily be looked up. Maybe this helps - or this :-)

p.s.: if you're somewhere near the road between Rødby and Helsingør, we might even be able to meet sometimes for a coffee and some C talk (C and coffee are somehow conceptionally connected).

adrianxw
Send message
Joined: 16 Aug 12
Posts: 26
Credit: 7,019,455
RAC: 10,362
Message 1702 - Posted: 20 Oct 2013, 9:02:21 UTC - in response to Message 1699.

SEL is/was Systems Engineering Laboratories of Fort Lauderdale, Florida., (info here) When I started, the directorate had just purchased a 32/77 computer running the MPX-32 operating system. For the first year, it was programmed in Fortran-66, until a F77 compiler became available. We, some years later, added a second SEL, a 32/27, which, despite the lower number, was perhaps more powerful then the 32/77, newer Silicon.

Another few years later, we bought 3 DEC VAX 11/750 mini's running VMS, and a few years later again, swapped them for 4 MicroVAX 3600's. Both configurations front ended by a host of PDP-11's. Real time pipe network SCADA systems, in F77.

The company was a big ICL mainframe customer, programmed in COBOL under the VME-B operating system, great for customer accounting etc. but useless for engineering and scientific work. Well, they probably could have done it, but the way they were set up was for online transaction processing, and a big mathematical model crunching a huge CPU bound task for 16 - 20 hours "dented" performance. The same job today could probably be done in 20 minutes on a mobile phone!

After leaving the firm, I worked for several, smaller companies, in C/C++ in the embedded market, (transport and safety systems mostly).

We live a short distance from junction 4 of the Helsingør motorway.

Profile Kevin
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 27 Jul 12
Posts: 507
Credit: 14,344,251
RAC: 1,727
Message 1703 - Posted: 20 Oct 2013, 23:11:35 UTC - in response to Message 1702.

That makes you nearly as old as me!
____________
Regards
Kevin
-----
International Centre for Radio Astronomy Research

adrianxw
Send message
Joined: 16 Aug 12
Posts: 26
Credit: 7,019,455
RAC: 10,362
Message 1707 - Posted: 23 Oct 2013, 15:26:25 UTC - in response to Message 1703.

What is left of me is 56.

Ananas
Send message
Joined: 13 Oct 12
Posts: 36
Credit: 2,500,031
RAC: 0
Message 1731 - Posted: 6 Nov 2013, 9:14:37 UTC - in response to Message 1707.

What is left of me is 56.

1958 built here and started with a TI 99/4A :-)

adrianxw
Send message
Joined: 16 Aug 12
Posts: 26
Credit: 7,019,455
RAC: 10,362
Message 1733 - Posted: 6 Nov 2013, 19:35:53 UTC - in response to Message 1731.

I didn't mention it above, but we had a few Intertec Superbrain micros early on in the office as well. I started micro programming with a breadboard NASCOM development system, no compiler or assembler, was programmed directly in hex. Used a Z80 microprocessor. I built a lighting control system for a nightclub with it.

Mumps [MM]
Send message
Joined: 9 Aug 12
Posts: 3
Credit: 101,965,316
RAC: 0
Message 1802 - Posted: 16 Nov 2013, 16:53:04 UTC - in response to Message 1643.

Let's see when (not if) we hit 50 TFLOPs! :)

Message boards : News : 37.1 TFlops


Main page · Your account · Message boards


Copyright © 2017 The International Centre for Radio Astronomy Research