Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
![]() |
World Community Grid Forums
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
No member browsing this thread |
Thread Status: Locked Total posts in this thread: 174
|
![]() |
Author |
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
A triple high today. We recorded the most points, results and CPU time ever.
|
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Having the BOINC database down put a severe crimp in our production. I'm glad it is back up.
Merry Christmas everyone! |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
We passed 100 million points today! Good work all!
|
||
|
David Autumns
Ace Cruncher UK Joined: Nov 16, 2004 Post Count: 11062 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Nice one D and crew
----------------------------------------![]() Superb Job ![]() |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I thought to my self, you're as happy about the daily strip,
as you are that no one stops the coffee import to germany, it's the least thing, to give a hello, when you join in here. So... Hello everybody! ![]() |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Welcome Delaz
|
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Thank you. Even if it is not much, I should be able to some 1.000 points
a month for the good cause, and the Team. The first 1.000 are made. ![]() |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Interesting.
I decided to do a comparison of my various machines. (Sadly, I can't edit my earlier post on the subject, but things have changed a little.) Doing just the basic 'CPU Benchmark' built into BOINC, I have compiled the following stats of average MIPS/MHz. (MIPS being an evenly weighted average of both Integer and Floating Point, MHz being just what it says. Processors with two cores have MHz doubled to account for their ability to do two full-speed calculations at once, processors with Hyperthreading are only counted once.) The winner, by far, is Core Duo. My Core Duo 2.0 GHz gets more than double the Integer MIPS of my Pentium Extreme Edition running at 4.0 GHz, per core. That means that ONE core of the Core Duo is faster than BOTH cores of the Pentium-EE when doing integer work. Core Duo: 2.905 MIPS/MHz Pentium M: 1.357 Celeron M: 1.271 G3: 1.054 G4 (average of two G4 computers:) 1.025 Pentium Extreme Edition: 0.976 Pentium 4: 0.600 That's right, a G4 is more efficient, running only one thread, than a dual-core Pentium 4, on a MIPS/MHz basis. Of course, nobody makes a 4 GHz G4, so that's a bit of a moot point. The best individual score is the Core Duo on Integer work, at 2.128 MIPS/MHz (per core,) the Pentium Extreme Edition, on the same Integer front, at 0.491 MIPS/MHz (per core.) At that rate, a QUAD-CORE based on Intel's (now dead) NetBurst core at 4.0 GHz would be beat by a single-core 2.0 GHz Core-based processor. The absolute worst score is the plain Pentium 4 on floating point, with 0.477 MIPS/MHz. (None of the machines beat 1.0 MIPS/MHz on floating point, while ONLY the NetBurst were under 1.0 for integer. The PowerPCs were all around 1.4-1.5 on integer, and the Celeron M and Pentium M (same basic core,) were at 1.7-1.8. (The changes in my lineup are that the 3.8 GHz Pentium 4 670 is now a 4.0 GHz (overclocked from 3.73 GHz,) Pentium Extreme Edition 965, and the 233 Mhz G3 has been retired completely. All other specs the same. So now my MHz total is up to 21,210 MHz, 5798 MB RAM. (Hrm, I miscalculated last time on the RAM.) And to think, someone overclocked one of Intel's new quad-core Xeons up to 3.0 GHz. One dual-socket system of that would be me in total GHz of processing power available! |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Eh. None of these metrics are really a good guide to performance. Raw clock speed in particular, is almost meaningless when compared over varying architectures.
But for WCG projects, I think the floating point performance is more significant than the integer performance. Certainly for the molecule modelling projects. The BOINC benchmark has flaws of its own, too (although it's based on dhrystone/whetstone). I don't want to discourage your numerical musings ($DEITY knows I've tried myself often enough) but you will have to consider many, many factors before you can draw firm conclusions. |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
crunchin&munchin for a cure!!!
![]() ![]() |
||
|
|
![]() |