| Author |
Message |
|---|
|
|
|
Link to the official thread
Just thought I'd start the thread, as AtP notice for this challenge. If anyone is new and/or in doubt, for LLR you should ideally turn hyperthreading OFF (via the BIOS) for LLR crunching.
If anyone has any tips/tricks, please fire away... |
|
|
|
|
|
Some benchmarks for an i7 6700k most welcome. Otherwise I will do some testing myself and post.
For an i7 3630QM (Ivy Bridge), I have 20.5 hrs per unit with 3 of 4 physical cores running based on some old notes. |
|
|
|
|
|
i7 6700k @4.4ghz ram@3000, ESP FFT length 864k ~27300sec |
|
|
|
|
i7 6700k @4.4ghz ram@3000, ESP FFT length 864k ~27300sec
Thanks, Steve. I expected a large difference relative to my i7 Ivy Bridge since this is the case for other subprojects as well. |
|
|
 Dave 
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
                           
|
|
I'm finishing this current b=49 sv then will do a 4 core benchmark. Nearly a year since I did any ESP (I checked my log). I'll be all-out with everything I've gt & reaching a holding point of 2.5M. |
|
|
|
|
i7 6700k @4.4ghz ram@3000, ESP FFT length 864k ~27300sec
4tasks, HT off, FFT length 960k ~32700sec |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
Count any Athlon64X2-5000s out of this. Looking like 200 hours a unit! |
|
|
|
|
|
I shall attend, but only at 50% effort. My 2600K box appears to be dead. The mobo has power (its little green LED is on) but the box won't fully power on. Maybe something mechanical with the "on" button. Dunno. Will work on it.
--Gary |
|
|
mackerelVolunteer tester
Send message
Joined: 2 Oct 08
Posts: 2645
ID: 29980
Credit: 568,565,361
RAC: 358
                           
|
|
If it is as simple as the power switch, you can replace with the reset switch if the case has one, or just use something conductive to momentarily short the power pins on the mobo. |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
i7-2600K 4 tasks 1600 RAM ~18h20. |
|
|
|
|
|
crap ola - my poor old systems will take a beating on this challenge. might get a few units in but not counting on doing much.
speaking of - that esp sieve project is really taking a long time to die - i kinda figure that the new ap27 is waiting on it to be removed before they add it since i haven't seen or heard of any new testing needing to be done. course our newest high priest might have more details. if not i am sure iain is causing the delay. |
|
|
|
|
crap ola - my poor old systems will take a beating on this challenge. might get a few units in but not counting on doing much.
speaking of - that esp sieve project is really taking a long time to die - i kinda figure that the new ap27 is waiting on it to be removed before they add it since i haven't seen or heard of any new testing needing to be done. course our newest high priest might have more details. if not i am sure iain is causing the delay.
One or two folks have been sitting on a handful of units and are responsible for the lack of progress. They must be doing the sieving manually with the Windows calculator. |
|
|
|
|
crap ola - my poor old systems will take a beating on this challenge. might get a few units in but not counting on doing much.
speaking of - that esp sieve project is really taking a long time to die - i kinda figure that the new ap27 is waiting on it to be removed before they add it since i haven't seen or heard of any new testing needing to be done. course our newest high priest might have more details. if not i am sure iain is causing the delay.
Yes!!!! Iain is the culprit!!!!!!
In all seriousness, I suspect he is working apace on getting production executables compiled. Which I in turn suspect is a far more critical dependency at this point than dropping a tombstone on esp-sieve.
|
|
|
|
|
Yes!!!! Iain is the culprit!!!!!!
In all seriousness, I suspect he is working apace on getting production executables compiled. Which I in turn suspect is a far more critical dependency at this point than dropping a tombstone on esp-sieve.
Actually there's not (too my knowledge) any connections between end of ESP-Sieve and start of AP27. The apps are essentially ready, we are just doing some final testing. I think it will go live soon(*).
(*) For some definition of soon.
- Iain
____________
Twitter: IainBethune
Proud member of team "Aggie The Pew". Go Aggie!
3073428256125*2^1290000-1 is Prime! |
|
|
|
|
Actually there's not (too my knowledge) any connections between end of ESP-Sieve and start of AP27. The apps are essentially ready, we are just doing some final testing. I think it will go live soon(*).
(*) For some definition of soon.
Another new badge to pursue. Yay! |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
On the subject of badge the badge will be AP26 but the project is AP27 ;). Anyone bothered by this? I'm not overly, just wonder if it's been pondered at an admin level. |
|
|
|
|
|
I was wondering about this. Maybe the whole thing should be changed to just an "AP" badge? I'm not fussed at all, it will be a new badge for me, and I'm happy about that. But I never participated at all in AP26, for example.
Also, one day I'll probably get a 321 Sieve badge, as more sieving will be required and the project is only "suspended". AP26 is "Retired".
Still, looking forward to earning another badge, whatever it says ;) |
|
|
|
|
|
Oops. I switched over to ESP-LLR in preparation for the challenge, but forgot to turn HT off in BIOS. I had it on due to all the recent sieving. Well, no huge loss; I'll let this batch finish (will take another day or so) then reboot.
On the bright side, I got my silver AP26/7 badge.
--Gary |
|
|
mackerelVolunteer tester
Send message
Joined: 2 Oct 08
Posts: 2645
ID: 29980
Credit: 568,565,361
RAC: 358
|
|
Doing some pre-testing now on what will be two of my fastest unit machines... it isn't looking good, at about 9 hours upwards.
Will be the first time I stress my mini-ITX build and I'm not looking forward to hearing that. I might even reduce the cooling on it, just so my ears don't bleed. |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
I have performed witchcraft + will be able to get a lowly extra 2 units for the team out my 2008 Athlon64X2-5000 (don't worry the rest will be all-out too). |
|
|
|
|
|
I will be all-in as well with 8 Skylake cores and another 3-4 Ivy Bridge cores ... all non-HT. |
|
|
|
|
|
Spinning down my gfn cpu tasks in preparation. All in, except my atom machines (they won't finish any units in time, they take over a day for a gfn-16).
|
|
|
|
|
|
TheDawgPound will be all in.
____________
There's someone in our head but it's not us. |
|
|
|
|
|
Will add my cores. Hope to crack top 500 |
|
|
|
|
|
I'm rolling! 10 WUs running within 30s of the start :)
- Iain
____________
Twitter: IainBethune
Proud member of team "Aggie The Pew". Go Aggie!
3073428256125*2^1290000-1 is Prime! |
|
|
|
|
I'm rolling! 10 WUs running within 30s of the start :)
- Iain
Nice. What ever happened to those monster "test" machines of yours? |
|
|
|
|
I'm rolling! 10 WUs running within 30s of the start :)
I played it safe by waiting 60-120 seconds with 11 physical cores running. I will wait for about three hours to pass before extracting runtimes. |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
I have 12 tasks going with the first at 15 secs. The Athlon64 maxing out @ 80C therefore maxing CPU fan needs to calm down. |
|
|
mackerelVolunteer tester
Send message
Joined: 2 Oct 08
Posts: 2645
ID: 29980
Credit: 568,565,361
RAC: 358
|
|
40 cores are go!
I can feal the heat already. The new build has a ton of coil whine from somewhere but thankfully contained by the case, and drowned out my my mini-ITX VR build which sounds like a swarm of angry wasps. I can't see that lasting the night before I do something about it...
Edit: I can't count. Make it 42 cores. |
|
|
|
|
|
Mackeral, respect with the 42 cores! |
|
|
|
|
|
wow oldman from ukraine has already posted a unit - if it's valid |
|
|
|
|
wow oldman from ukraine has already posted a unit - if it's valid
I want to see the runtimes from fastgeek or whatever his alias is. |
|
|
|
|
wow oldman from ukraine has already posted a unit - if it's valid
And just like that it's gone from the stats page ....
____________
There's someone in our head but it's not us. |
|
|
|
|
|
Fellow rodents, lend me your ears.
Something very strange is going on with about 60 minutes worth of processing.
The i3 6100 (3.7 GHz) and i3 6300 (3.8 GHz) are showing about a 10-hr runtime per unit per core (both physical cores running) and which is what I expected before the start of the challenge.
The i7 6700k (4.0 GHz) is showing about a 20-hr runtime per unit per core with all four physical cores running.
My i7 3630QM (3.2 GHz) is showing around 18.5 hrs runtime per unit per core with three of the four physical cores running (the other physical core is dedicated to PPR12M and Collatz).
No HT with CPU usage set at 50% for all of the Skylake systems.
I am stomped by the performance of the i7 6700k.
Thoughts? |
|
|
|
|
wow oldman from ukraine has already posted a unit - if it's valid
And just like that it's gone from the stats page ....
Hmmm a magic trick that made itself vanish |
|
|
|
|
|
TheDawgz believe that these work units vary considerable in the number of iterations needed based on the "n" involved.
At this moment in time, the difference between the smallest and largest "n" is ~300,000
Remember that we are looking at some # * 2 ^ n +1 with the "n" varying by 300,000
TheDawgz could be terribly wrong about this. We hate math as much as we hate cats.
____________
There's someone in our head but it's not us. |
|
|
|
|
TheDawgz could be terribly wrong about this. We hate math as much as we hate cats.
We had a cat once. It had a hat |
|
|
|
|
TheDawgz believe that these work units vary considerable in the number of iterations needed based on the "n" involved.
At this moment in time, the difference between the smallest and largest "n" is ~300,000
Remember that we are looking at some # * 2 ^ n +1 with the "n" varying by 300,000
TheDawgz could be terribly wrong about this. We hate math as much as we hate cats.
This was the first thought which sprung to mind ... I am truly puzzled. |
|
|
|
|
TheDawgz could be terribly wrong about this. We hate math as much as we hate cats.
As much as I LOVE canines, I also have a dislike for felines. |
|
|
|
|
Nice. What ever happened to those monster "test" machines of yours?
I had a prod at it, but since my 'testing' a few months ago they've locked down the compute nodes so they don't have direct access to the outside network. It might be possible to set up a tunnel, but not straightforward...
- Iain
____________
Twitter: IainBethune
Proud member of team "Aggie The Pew". Go Aggie!
3073428256125*2^1290000-1 is Prime! |
|
|
|
|
|
All of my computers are crated and in transit, but I spun up a flock just for the Challenge. We'll see what they can squeeze out.
____________
Proud member of Team Aggie the Pew
"Wir müssen wissen. Wir werden wissen."
"We must know, we shall know."
- David Hilbert, 1930 |
|
|
|
|
I am stomped by the performance of the i7 6700k.
Problem solved: according to CPU-Z, the RAM is single channel. Someone is going to get a mouthful (dealer for the RAM). |
|
|
|
|
|
As an aside, it's sometimes more efficient (in the long run) to leave a core free as well as HT off, with LLR.
I say sometimes: For evidence, numbers would have to be produced pre-every challenge to prove it, and of course if the challenge is only a couple of days long, then it's probably best to go all-out, especially if you can micro-manage when the last batch is about to start.
In under 5 hours from now, the challenge scoreboard is going to fill up all at once. |
|
|
|
|
|
Nice job Steve! On the board in first report. |
|
|
|
|
|
And mack is on the board as well! |
|
|
|
|
I am stomped by the performance of the i7 6700k.
Problem solved: according to CPU-Z, the RAM is single channel. Someone is going to get a mouthful (dealer for the RAM).
I'm thinking the dimms are in the wrong slots. Some motherboards are cranky about where they need to go.
|
|
|
|
|
I am stomped by the performance of the i7 6700k.
Problem solved: according to CPU-Z, the RAM is single channel. Someone is going to get a mouthful (dealer for the RAM).
I'm thinking the dimms are in the wrong slots. Some motherboards are cranky about where they need to go.
Zimmer,
I cannot thank you enough! It was indeed a case of a misplaced stick (slots 4 and 3). I just did the change to slots 4 and 2 and CPU-Z is now reporting dual channel and the newly downloaded units are showing the expected runtime during the first five minutes of execution.
I owe you a beer!!! |
|
|
|
|
Nice job Steve! On the board in first report.
http://www.primegrid.com/result.php?resultid=738209017
FFT length only 864k.
system: i7 6700k @4.4ghz ram@3100 (it's a 3200 ram kit, but had no time to tweak/bench it, so running on "save" LLR settings)
sad that I only have this system with 4 tasks running and 2 slower skylake cores ;)
btw. fastgeek is hitting hard:
http://www.primegrid.com/hosts_user.php?userid=371522
I'm counting 886 cores with ESP tasks running. Half of the cores should be HT ... but still crazy. |
|
|
|
|
I owe you a beer!!!
Just happy to help-but I will never refuse a beer. |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
btw. fastgeek is hitting hard:
http://www.primegrid.com/hosts_user.php?userid=371522
I'm counting 886 cores with ESP tasks running. Half of the cores should be HT ... but still crazy.
I don't see him yet - are they all going to start landing at once?
Is there going to be a zune v fastgeek battle? |
|
|
|
|
I don't see him yet - are they all going to start landing at once?
They will all hit in a small timeframe.
2.4-2.6ghz skylake runtime should be around 18-22h for one physical core. With HT on, time should be about *2. But it looks like that for this large FFT tasks the RAM speed and channel configuration is more important. |
|
|
|
|
|
I know Michael and possibly all the mods won't like this too much, but I'm not sending out a mass public message or anything:
If you have few computers, or even a single one, but they're\it's fast - when your running tasks are nearing completion, it might be worth aborting the ones which have been waiting to start for about 10 hours in a challenge situation, and then updating BOINC, because the new ones which will replace those tasks will be larger. This probably won't be true in a non-challenge situation, because tasks won't get munched up anywhere near as fast.
That is the theory, and certainly I just experimented and the "Numbers" in the name of the tasks waiting to start just increased. I have no idea whether this means that they'll have higher FFTs and yield more credit or not. I don't want to assume, it's just an idea. Of course, larger tasks will take a bit longer to process, but in a 5-day challenge where tasks last about 12 hours, it's probably worth the time and credit, by the end. At least, it might nudge your machine a place or three above another machine which has been left to crunch blindly, for example. |
|
|
|
|
|
With 24 hours done and 96 to go ....
Teams ( out of 41 with credit )
1 Aggie The Pew 1,229,556.95
2 Czech National Team 1,096,141.72
3 Sicituradastra. 929,747.40
4 SETI.Germany 806,984.11
5 Crunching@EVGA 625,260.79
Individuals ( out of 141 with credit )
2 TheDawgz 386,158.18
8 zunewantan 209,508.61
9 mackerel 203,328.49
12 Van Zimmerman 123,835.35
19 mattozan 93,084.94
38 Steve 46,059.91
41 Iain Bethune 44,645.35
54 Anthony Ayiomamitis 36,296.14
73 Dave 23,076.33
78 brinktastee 18,677.73
119 Dirk 13,413.72
121 gazzyk1ns 12,875.43
127 1998golfer 9,321.05
128 Rick Reynolds 9,275.71
Go Rats! Go!!
____________
There's someone in our head but it's not us. |
|
|
|
|
|
Thanks, Dawgz.
I lost the first 12 hrs for nothing due to the "single channel" RAM monkey business and where I eventually aborted everything. Nevertheless, I am now back in full production mode and have already moved up.
Eleven cores hacking away! |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
The i3-4120 just dumped - 1d1h45 duration 4 tasks (HT is on & I can't get into BIOS as it's remote but it normally spends its time sieving). |
|
|
|
|
|
ESP-LLR: Summer Paralympics Challenge (2016-09-02 18:00:00 to 2016-09-07 18:00:00)
Last update: 2016-09-04 16:30:02
Rank Name Score
1 Czech National Team 3571626.31
2 Aggie The Pew 3283909.48
Yep, the CNT did it again...
Added 16 cores , hope this helps (a bit)
Go Aggie! |
|
|
|
|
|
Given how close we are, I'm launching a few birds (code name "chickenhawk"). Since I'm also lazy, and already have an account with digitalocean, I'm giving them a whirl. So far they are running about 4% per hour per core on ESP. |
|
|
|
|
|
With 48 hours done and 72 to go ....
Teams ( out of 61 with credit )
1 Czech National Team 3,757,869.07 (40 users with challenge credit)
2 Aggie The Pew 3,487,328.42 (16 users with challenge credit)
3 Sicituradastra. 2,934,841.41 (16 users with challenge credit)
4 SETI.Germany 2,186,722.82 (18 users with challenge credit)
5 Crunching@EVGA 1,691,414.13 (11 users with challenge credit)
Individuals ( out of 254 with credit )
2 TheDawgz 961,482.38
4 zunewantan 817,497.40
9 mackerel 559,403.06
14 Van Zimmerman 297,052.00
31 mattozan 187,504.75
37 Anthony Ayiomamitis 144,911.45
53 Steve 107,522.96
62 Iain Bethune 95,911.90
95 Dave 65,047.25
105 brinktastee 56,425.17
120 Dirk 55,049.86
129 Walter Darimont 42,212.68
134 gazzyk1ns 41,033.98
188 Gary Craig 18,817.01
194 1998golfer 18,736.15
195 Rick Reynolds 18,720.43
Go Rats! Go!!
____________
There's someone in our head but it's not us. |
|
|
|
|
Thanks, Dawgz.
I lost the first 12 hrs for nothing due to the "single channel" RAM monkey business and where I eventually aborted everything. Nevertheless, I am now back in full production mode and have already moved up.
Eleven cores hacking away!
Its amazing how much just a single 6700k firing on all cylinders matters in those rankings. |
|
|
|
|
Thanks, Dawgz.
I lost the first 12 hrs for nothing due to the "single channel" RAM monkey business and where I eventually aborted everything. Nevertheless, I am now back in full production mode and have already moved up.
Eleven cores hacking away!
Its amazing how much just a single 6700k firing on all cylinders matters in those rankings.
I agree fully and as illustrated in the table of results from Dawgz.
My i7 3630QM will be dumping results twice over the next 24 hrs which will help further in my climb.
Nice to finally be making a meaningful contribution. |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
Fantastic efforts all round, still full bore this end with what I have at my disposal. One can never have enough. |
|
|
|
|
|
fastgeek's first WU round is coming now slowly .... runtime >40h, but with so many cores it will get a mess.
It's really interesting how good an overclocked i7 6700k is in ESP. Nearly 100k credits with this machine only. |
|
|
|
|
|
I won't be much help on this one. My wife and I got in the car and drove to Santa Fe just to get away. Spur of the moment thing. Just pulled in. 16 hours in two days. It's nice that most of the west states raised the speed limit to 70-75. Got behind one guy doing 90 for a while.
Now it's time to be a tourist!
____________
|
|
|
TylerProject administrator
Volunteer tester
Send message
Joined: 4 Dec 12
Posts: 1078
ID: 183129
Credit: 1,378,716,398
RAC: 37,605
             
|
|
Wow... I guess my setup is getting outdated after all, I'm disappointed in these 16 hour runtimes, I used to be able to keep up with most of the quick machines.. Sigh. i5-2500k@4.4GHz, ram 1866 dual channel, dual rank. Thought I'd be faster than this, oh well. Have some money leftover from school... Could save it, or upgrade my setup... Hmm...
____________
275*2^3585539+1 is prime!!! (1079358 digits)
 Proud member of Aggie the Pew
 |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
At the halfway point not sure when to switch to 3 units. 4 is best throughput - maybe 3 units on the last batch. Or just wing it. |
|
|
|
|
At the halfway point not sure when to switch to 3 units. 4 is best throughput - maybe 3 units on the last batch. Or just wing it.
I will be spending an extra 10 days with ESP LLR post-challenge so as to get up to Ruby. As a result, micro-management on the last batch for the purposes of the challenge and then back to normal processing for the final leg towards Ruby.
For each verified unit, I have another two pending ... it goes to show how dominant the Skylake CPU's are with ESP LLR units (and presumably LLR in general). I am delighted with my decision to go Skylake (i3 6100 3.7 GHz, i3 6300 3.8 GHz, i7 6700k 4.0 GHz) and any possibly new systems forthcoming will be exclusively Skylake. |
|
|
mackerelVolunteer tester
Send message
Joined: 2 Oct 08
Posts: 2645
ID: 29980
Credit: 568,565,361
RAC: 358
|
|
Skylake is the fastest per clock, provided you can feed it enough ram bandwidth. Previously testing showed it has about 14% IPC over Haswell. Put in slow ram, it'll be crippled like any other CPU. My Haswell i3 and Broadwell i5 are competitive per core with upper-mid range Skylake.
I think right now it doesn't make much sense to buy anything other than Skylake, with the possible exception of Broadwell (non "-E") if the price is right, and only because of the L4 cache on that negating the need for performance ram. |
|
|
|
|
fastgeek's first WU round is coming now slowly .... runtime >40h, but with so many cores it will get a mess.
Out of the blue he is now #2 overall. Tremendous volume at play ... albeit at a slow pace.
Edit: Some work units took 65 hrs to complete. That means only one batch by the deadline unless he cuts down on the volume. |
|
|
|
|
|
With 72 hours done and 48 to go ....
Teams ( out of 68 with credit )
1 Czech National Team 6,731,563.53 (42 participants)
2 Aggie The Pew 6,035,345.85 (16 participants)
3 Sicituradastra. 5,103,154.14 (18 participants)
4 SETI.Germany 3,698,967.86 (25 participants)
5 Crunching@EVGA 2,831,794.87 (11 participants)
Individuals ( out of 326 with credit )
4 TheDawgz 1,579,166.60
5 zunewantan 1,451,920.98
9 mackerel 911,509.67
14 Van Zimmerman 592,127.13
34 mattozan 301,614.00
40 Anthony Ayiomamitis 249,786.41
63 Iain Bethune 167,166.43
73 Steve 155,069.69
86 Walter Darimont 131,281.72
102 Dave 112,471.91
118 brinktastee 94,494.01
129 Dirk 88,240.65
151 gazzyk1ns 69,512.54
178 Rick Reynolds 55,570.01
207 1998golfer 37,747.82
208 Gary Craig 37,666.29
Go Rats! Go!!
____________
There's someone in our head but it's not us. |
|
|
|
|
|
I bet Z is smarter than your average mass-cruncher. The end is going to be interesting :big smiley face: |
|
|
TylerProject administrator
Volunteer tester
Send message
Joined: 4 Dec 12
Posts: 1078
ID: 183129
Credit: 1,378,716,398
RAC: 37,605
|
|
Wow... Just realized I'm dumb and crippling my crunching performance. My second stick of RAM is sitting on my desk, while the other one is in my computer... I took it out to do something that required a max of 4GB of ram, didn't think to put it back in... Oops. Fixing that now..
____________
275*2^3585539+1 is prime!!! (1079358 digits)
Proud member of Aggie the Pew
|
|
|
|
|
|
Aargh, a ups tripped offline, and I lost a couple of machines for the (long) weekend.
|
|
|
|
|
Aargh, a ups tripped offline, and I lost a couple of machines for the (long) weekend.
I had a 40-min power failure about 90 minutes ago thanks to a serious downpour. All systems now back up and running again but three tasks will now just miss the 21:00 deadline this evening by less than 15 minutes as a result and which means I may need some very minor micromanagement tomorrow also at 21:00 when the challenge comes to an end since the average runtime on my i7 6700k is 11.78 hrs/task/core. |
|
|
|
|
...average runtime on my i7 6700k is 11.78 hrs/task/core.
You running 4 cores? I'm looking at 9.14 hours/task/core with 3. |
|
|
|
|
...average runtime on my i7 6700k is 11.78 hrs/task/core.
You running 4 cores? I'm looking at 9.14 hours/task/core with 3.
Yes, 4 cores at 4.0 GHz. |
|
|
|
|
|
With 96 hours done and 24 to go ....
Teams ( out of 74 with credit )
1 Czech National Team 9,734,693.61 (46 participants)
2 Aggie The Pew 8,708,279.96 (16 participants)
3 Sicituradastra. 7,288,576.48 (18 participants)
4 SETI.Germany 5,612,719.23 (27 participants)
5 [H]ard|OCP 4,215,435.28 (5 participants)
Individuals ( out of 376 with credit )
3 zunewantan 2,273,340.36
5 TheDawgz 2,134,720.20
10 mackerel 1,232,080.88
14 Van Zimmerman 922,144.73
30 mattozan 459,708.47
43 Anthony Ayiomamitis 335,938.79
70 Steve 231,688.76
77 Iain Bethune 214,987.98
91 Walter Darimont 188,296.74
97 Dave 169,935.64
115 brinktastee 137,734.68
130 Dirk 126,397.66
167 gazzyk1ns 88,664.75
175 Gary Craig 76,005.28
193 Rick Reynolds 69,310.68
234 1998golfer 47,324.37
Go Rats! Go!!
____________
There's someone in our head but it's not us. |
|
|
|
|
|
I'll be turning in another 20 units before the deadline, but that's all for the flock.
____________
Proud member of Team Aggie the Pew
"Wir müssen wissen. Wir werden wissen."
"We must know, we shall know."
- David Hilbert, 1930 |
|
|
mackerelVolunteer tester
Send message
Joined: 2 Oct 08
Posts: 2645
ID: 29980
Credit: 568,565,361
RAC: 358
|
|
It's 32C in my front room now. This challenge can't end soon enough! My Xeon box is already NNW, and I'll set the rest of the systems tomorrow morning.
I hope in my using of one system to test SSDs earlier I haven't taken too long that it misses getting its last units out in time... gonna be really tight now but only 2 units "lost" even if so. |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
The 2008 Athlon64X2 successfully completed its 2 last night. Trying to squeeze in 2 from the i3 but saying 15 hrs with 13 hrs left of challenge so might not make it. The Celeron N is going to take 116 hrs as predicted so will make it. |
|
|
mackerelVolunteer tester
Send message
Joined: 2 Oct 08
Posts: 2645
ID: 29980
Credit: 568,565,361
RAC: 358
|
|
With just under 10 hours to go, anything not started isn't going to finish in time so NNW set around now. Hope it will be a lot cooler by the time I get home in the evening... |
|
|
|
|
|
Have 2 units left.. should get turned in with plenty of time to spare. |
|
|
|
|
|
Looks like I'm going to be down to the wire! Got another set of 6 that should be finished within about 4 hours (4:12 left in the challenge clock as I write), and 2 more that are running on my quad-core (2 cores idle to speed them up), that might just sneak in...
- Iain
____________
Twitter: IainBethune
Proud member of team "Aggie The Pew". Go Aggie!
3073428256125*2^1290000-1 is Prime! |
|
|
mackerelVolunteer tester
Send message
Joined: 2 Oct 08
Posts: 2645
ID: 29980
Credit: 568,565,361
RAC: 358
|
|
7 units left here... may my house be much cooler this evening.
Wondering if I can hold 10th place individually. #11 reduced gap by around 50k in last two hours, hope they don't have more to dump in remaining time. |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
1 unit left, i3 on just 1 core - should snick in within the final quarter-hour. |
|
|
|
|
|
27 left, I doubt more than 9 will make it in before the deadline. |
|
|
|
|
|
24 left of which 3 will be completed before deadline...
(quite some are only approx hour short..)
Go Aggie !!! GOOoooo! |
|
|
|
|
|
Dozed off on the couch ... 90 minutes lost due to yet another power failure as a result of bad weather conditions the past 24 hours. Battling to get 2 or 3 units to make the deadline when all four from the same system should have been a breeze prior to the two power failures during the past 24 hrs. |
|
|
|
|
|
Rest assured, I'm here to save the day - I've timed my last ONE WU almost perfectly, so it'll hit before the deadline. No need to thank me. |
|
|
mackerelVolunteer tester
Send message
Joined: 2 Oct 08
Posts: 2645
ID: 29980
Credit: 568,565,361
RAC: 358
|
|
I'm done with all my units in. About 4 WUs could swing me down a place so let's see what the next 40 minutes bring... |
|
|
|
|
I bet Z is smarter than your average mass-cruncher. The end is going to be interesting :big smiley face:
Gazzy, you were right ! BOOM-Z has struck again:
ESP-LLR: Summer Paralympics Challenge (2016-09-02 18:00:00 to 2016-09-07 18:00:00)
Last update: 2016-09-07 17:30:03
Rank Name Score
1 Aggie The Pew 13981942.26
2 Czech National Team 13458723.77 |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
I'm done. + wowsers z-san! He knows the technique ;). |
|
|
|
|
I bet Z is smarter than your average mass-cruncher. The end is going to be interesting :big smiley face:
Gazzy, you were right ! BOOM-Z has struck again:
ESP-LLR: Summer Paralympics Challenge (2016-09-02 18:00:00 to 2016-09-07 18:00:00)
Last update: 2016-09-07 17:30:03
Rank Name Score
1 Aggie The Pew 13981942.26
2 Czech National Team 13458723.77
I was about to post a message to this effect ... boy did he drop a huge Z-bomb!!!
He jumped from 4th or so to 1st place overall. Fastgeek will not be believing his eyes.
Edit: My pending credit dropped from 195k to 157k ... thanks to Z-san? :-) |
|
|
|
|
|
The UNOFFICIAL final standings are .... ....
Teams ( out of 78 with credit )
1 Aggie The Pew 14,062,884.96 (16 participants)
2 Czech National Team 13,539,452.58 (49 participants)
3 Sicituradastra. 9,637,600.62 (19 participants)
4 SETI.Germany 8,436,370.18 (31 participants)
5 [H]ard|OCP 6,330,904.61 (5 participants)
Individuals ( out of 424 with credit )
1 zunewantan 5,702,082.02
5 TheDawgz 2,734,160.04
10 mackerel 1,543,315.68
15 Van Zimmerman 1,270,976.52
33 mattozan 564,232.60
46 Anthony Ayiomamitis 422,253.46
72 Walter Darimont 296,053.60
73 Iain Bethune 291,839.07
80 Steve 279,574.59
102 Dave 241,476.62
136 brinktastee 170,510.35
146 Dirk 155,194.87
169 gazzyk1ns 126,452.30
186 Rick Reynolds 107,633.21
195 Gary Craig 95,331.41
244 1998golfer 61,798.61
CongRATs to all the RATS !!
____________
There's someone in our head but it's not us. |
|
|
|
|
|
I forgot to report WUs and was going to lose if I did not wake up at AM 2:00
CongRATs!!
Z |
|
|
|
|
I forgot to report WUs and was going to lose if I did not wake up at AM 2:00
CongRATs!!
Z
I was just sitting here in amazement! What a job!
Well DONE AtP !!! I bet there are a ton of surprised folks and teams out there. We however know what can happen !!!
Congrats to Z for waking up or staying up and for the team for a wonderful performance!
Cheers |
|
|
|
|
I forgot to report WUs and was going to lose if I did not wake up at AM 2:00
CongRATs!!
Z
That was very amusingly done from a rat point of view Z-san. My sincere thanks. I am not entirely sure the Czechs will enjoy the joke as much as us. |
|
|
|
|
|
I do believe Z-San does not like to lose :) |
|
|
|
|
I forgot to report WUs and was going to lose if I did not wake up at AM 2:00
CongRATs!!
Z
That was very amusingly done from a rat point of view Z-san. My sincere thanks. I am not entirely sure the Czechs will enjoy the joke as much as us.
Only the Czechs? Fastgeek will change his alias to WTFgeek! LOL. |
|
|
|
|
I do believe Z-San does not like to lose :)
TheDawgz concur!!
Thank you Z !!
____________
There's someone in our head but it's not us. |
|
|
|
|
|
These rats sure know how to scurry at the end! CongRATulations to zunewantan and Aggie the Pew!
My last WU missed the deadline by about 20 minutes, but it all went pretty smoothly by remote control from my vacation hideout.
Was anyone running AWS instances? If so, I'd be curious to know what your runtimes were? The Flock was averaging about 62,000 seconds (MIN 56,600 MAX 74,700 MED 61,000 n=120)
____________
Proud member of Team Aggie the Pew
"Wir müssen wissen. Wir werden wissen."
"We must know, we shall know."
- David Hilbert, 1930
|
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
Fantastic achievement all round - love that 24/7 dedication Z-san! That's what makes it intense - amazing how awake you can be at stupid o'clock ;). |
|
|
|
|
These rats sure know how to scurry at the end! CongRATulations to zunewantan and Aggie the Pew!
My last WU missed the deadline by about 20 minutes, but it all went pretty smoothly by remote control from my vacation hideout.
Was anyone running AWS instances? If so, I'd be curious to know what your runtimes were? The Flock was averaging about 62,000 seconds (MIN 56,600 MAX 74,700 MED 61,000 n=120)
I looked at the final standings and started laughing. I wonder if Z-san actually calculates "I need to drop this many units in the last hour..." And then times it perfectly.
I was running digitalocean instances, and the best times I got were 67000 seconds in either their San Fran 2 data center or their London data center. Most others were in the 76-82000 second range. New York 1 and Amsterdam were terrible, with 110-130000 second run times.
|
|
|
|
|
Congrats to Z for waking up or staying up and for the team for a wonderful performance!
Cheers
Same from me! He had me thinking he might not have had it this time, too, towards the end! |
|
|
|
|
|
I missed the deadline with my last two units (not that it would have made the difference anyway...). Great team showing though :)
- Iain
____________
Twitter: IainBethune
Proud member of team "Aggie The Pew". Go Aggie!
3073428256125*2^1290000-1 is Prime! |
|
|
|
|
|
With two WU's I'm missing the deadline by approximately 9 days. I must have made a mistake somewhere.
A good thing that Z-san timed his coup de théatre to perfection.
What a guy to have in our team!
And congratulations for the entire team. |
|
|
|
|
I forgot to report WUs and was going to lose if I did not wake up at AM 2:00
CongRATs!!
Z
That was very amusingly done from a rat point of view Z-san. My sincere thanks. I am not entirely sure the Czechs will enjoy the joke as much as us.
Only the Czechs? Fastgeek will change his alias to WTFgeek! LOL.
http://www.primegrid.com/forum_thread.php?id=6958&nowrap=true#98677 ... all the marbles! |
|
|
|
|
|
Looks like the glove has been thrown |
|
|
|
|
Looks like the glove has been thrown
I think Z-san may change his name to Niten Dōraku. |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3206
ID: 130544
Credit: 2,284,240,700
RAC: 922,232
|
|
He will demonstrate Moves that The Old Masters have never seen >.< |
|
|
|
|
Looks like the glove has been thrown
I think Z-san may change his name to Niten Dōraku.
I still insist on WTFgeek. :-)
It is sad when you underestimate your very powerful opponent. |
|
|
|
|
|
I've moved up a place, presumably because someone else produced erroneous results.
That is all. |
|
|
