Posts by [AF>Amis des Lapins] Phil1966

1) (Message 7780)
Posted 11 Mar 2023 by Profile [AF>Amis des Lapins] Phil1966
Post:
@Kyong
2) (Message 5375)
Posted 21 Apr 2017 by Profile [AF>Amis des Lapins] Phil1966
Post:
:(
3) (Message 5371)
Posted 17 Apr 2017 by Profile [AF>Amis des Lapins] Phil1966
Post:
Hello,
Would it be possible to prevent "Ghost Crunchers / computers" to upload
a full cache of WU's (sometimes > 1000) without calculating them.
The server already cancelled some of these uncalculated WU's, will cancel the remaining WU's automaticaly, but these crunchers will again and again
receive a brand new batch of WU's that they won't return in time.
As I was wondering why I have so many "waiting validation" (> 500),
I found at least 3 or 4 computers having > 1000 WU's + 0 returned WU
since April 11th and already several series of "cancelled by the server" WU's.
Thank You
Best Regards
4) (Message 5340)
Posted 9 Apr 2017 by Profile [AF>Amis des Lapins] Phil1966
Post:
If you need to buy a biger server, don't hesitate to ask for a BITCOIN UTOPIA support ;)
5) (Message 5330)
Posted 9 Apr 2017 by Profile [AF>Amis des Lapins] Phil1966
Post:
Thank you
6) (Message 3835)
Posted 30 Nov 2014 by Profile [AF>Amis des Lapins] Phil1966
Post:
Hello,

could someone explain to me why there are so many changes in the WU's lenght ?

Below instances = AVX, (all on the same machine - ID 54940 !) but same happens with SSE2 and SSE3 : ALL WU's need > 2x more time to complete than 4 or 5 days ago ... ?

Thank You



54594862 23583313 24 Nov 2014, 18:50:57 UTC 26 Nov 2014, 6:42:01 UTC Terminé et validé 2,752.98 2,725.56 480.00 Period Search Application

54594807 23583289 24 Nov 2014, 18:50:57 UTC 26 Nov 2014, 6:42:01 UTC Terminé et validé 2,808.27 2,779.88 480.00 Period Search Application

54589306 23580566 24 Nov 2014, 18:28:10 UTC 26 Nov 2014, 6:42:01 UTC Terminé et validé 2,951.73 2,922.96 480.00 Period Search Application




56074104 24296773 29 Nov 2014, 20:03:37 UTC 30 Nov 2014, 2:17:07 UTC Terminé et validé 7,282.37 7,203.83 480.00 Period Search Application

56074360 24296907 29 Nov 2014, 20:03:37 UTC 30 Nov 2014, 0:15:43 UTC Terminé et validé 7,424.55 7,312.80 480.00 Period Search Application

56074210 24296854 29 Nov 2014, 20:03:37 UTC 30 Nov 2014, 4:21:07 UTC Terminé et validé 7,585.05 7,400.72 480.00 Period Search Application
7) (Message 3811)
Posted 23 Nov 2014 by Profile [AF>Amis des Lapins] Phil1966
Post:
Hi Mikey,

Thank you for your message.

Sorry, but ...

a) even if a dedicate a cpu core / GPU WU, (app_config) it slows down and sometimes freezes the CPU WU's ???

b) CPU time shown in the stats don't confirm what you are writting.

Less than 10 seconds CPU time per GPU WU (CUDA55)

c) ie on GPUGRID, I know how to "oblige" BM to use the core CPU it is supposed to use, and there is a real gain. (CPU time = GPU time)

d) don't understand why, but when running Asteroids on CPU + GPU, my machines are crashing

e) What's the point wasting GPU time when the gain is only 50 % ag AVX WU's ?
If optimzed, these CUDA app should/could run in about 10 minutes ...

Nevertherless, I like the project, and I try to understand how to improve crunching on it :)

Best

Phil1966
8) (Message 3798)
Posted 20 Nov 2014 by Profile [AF>Amis des Lapins] Phil1966
Post:
ASTEROIDS GPU is not only slowing the CPU very seriously,
but today I tried to run it on CPU + GPU.

All CPU WU's are in "waiting mode".

Waiting for what, I don't know.

As far as I know, it's the only GPU project having such a negative impact on the CPU while it is not supposed to.

Even if you dedicate a CPU core to ASTEROIDS GPU.

It looks like no improvement was made on these GPU apps since launch,
regardless the numerous crunchers remarks.

It's "sad".

Best,

Ph.
9) (Message 2775)
Posted 16 Mar 2014 by Profile [AF>Amis des Lapins] Phil1966
Post:
Not sure why you want an app_info but then I'm app_info challenged.

If it helps... I'm getting the AVX app on my Haswell system without an app_info. All I did was plug it in and turn it on. HA_Soft says it's because Linux detects AVX properly but Windows can't find its own ass with both hands and a mirror. As usual.

You will get a mix of SSE2 and AVX tasks for a while until the server detects that your system does AVX way faster than SSE2 then from that point on you'll receive only AVX.

As for crunching here with your GPUs, unless you have Titans or Teslas or something with decent DP float power you'll find tasks run faster on Haswell than on GPU, I think. I tried my 670 here a while back and the times were not what I would call impressive and it's because DP is crippled on GTX 670. They have improved the GPU app at least once since then but if I understand correctly it's still slower than AVX on anything other than Titan and Tesla which have full (uncrippled) DP float capability.

Still, I didn't need an app_info for GPU either, just had to turn GPU on in website prefs.


You're 100 % right, except that I don't think Asteroids uses DP. (see this post : http://asteroidsathome.net/boinc/forum_thread.php?id=268)
Like your point of vue about Windows :D
Ran some tests yesterday, and AVX "better" than GPU.
http://asteroidsathome.net/boinc/forum_thread.php?id=268
10) (Message 2765)
Posted 15 Mar 2014 by Profile [AF>Amis des Lapins] Phil1966
Post:
Hello !

As discussed here : http://asteroidsathome.net/boinc/forum_thread.php?id=200&postid=2758#2758

I made some tests today, and here are the results + questions.

CPU tests

Crunchbox = ASUS Z87 Pro / i7 4770K @ 3.9 Ghz / AIO WC Nepton 140XL / etc

i7 4770K @ 3.9 Ghz - HT OFF - 3 AVX WU's at the same time = Aver. 3677 secs / WU = 11040 PPD / CORE
i7 4770K @ 3.9 Ghz - HT ON*** - 7 AVX WU's at the same time : Aver. 5460 secs / WU = 7680 PPD / CORE
i7 4470K @ 3.9 Ghz - HT ON - 7 SSE3 WU's at the same time = Aver. 8100 secs / WU = 5120 PPD / CORE

*** A powerfull Aircooling or Watercooling is MANDATORY to run AVX WU's with HT ON !

=> One can see the benefits of each type of WU's

Thank you for the various applications you offer. It would be nice to have an option in the account preferences rather to be obliged to use an app_info.xml file. (Not user friendly for beginners)


GPU tests

GTX660Ti CUDA 55 : 3767 secs / WU = 11040 PPD = same as 1 CORE CPU AVX HT OFF

GTX770 : 3450 secs / WU = 12000 PPD = little better than 1 CORE CPU AVX HT OFF

=> How comes the GPU tasks are so slow, given the fact that :

All WU's = 1200020 GFLOPS
A much more significant improvement was observed on other projects that propose
to crunch on GPU or CPU the same type of WU's
ie : PPS Sieve & GFN Short : GPU WU's > 15 times faster than CPU.

As I am not a specialist, and with all due respect, I just want to understand why the same kind of speed cannot be reached with Asteroids.

Finally, once running a GPU task, the crunchbox slows down + ALL (CPU+GPU) WU's disappear when closing BM.

Thank You

Kind Regards,

Phil1966
11) (Message 2761)
Posted 15 Mar 2014 by Profile [AF>Amis des Lapins] Phil1966
Post:
Meanwhile we have increased wu's lenght. Now it's about 3500s.


Understood !

Thank You for your quick answer ;)

Kind regards

Philippe

NB I am still planning to run a test on my crunchbox as soon as the project becomes available :D
12) (Message 2760)
Posted 15 Mar 2014 by Profile [AF>Amis des Lapins] Phil1966
Post:
My new AMD A10-7850K with Steamroller cores seems to be so far a little faster with the AVX version. Unfortunately, he gets currently only SSE2 WUs.
Perhaps this will change when the whole pending jobs are valid.
SSE2/3: ~ 9300 s
AVX: ~ 8900 s


Hello !

If you run XP/W7/W8 you need an app_info.xml in order to receive AVX WU's.

ie (this is the one I found on the forums and use with W7 / i7 4770K / HT OFF)

Philippe


<app_info>
<app>
<name>period_search</name>
<user_friendly_name>Asteroids</user_friendly_name>
</app>

<!-- AVX -->
<file_info>
<name>period_search_10210_windows_x86_64__avx.exe</name>
<executable/>
</file_info>
<app_version>
<app_name>period_search</app_name>
<version_num>1021</version_num>
<avg_ncpus>1.00</avg_ncpus>
<max_ncpus>1.00</max_ncpus>
<plan_class>avx</plan_class>
<file_ref>
<file_name>period_search_10210_windows_x86_64__avx.exe</file_name>
<main_program/>
</file_ref>
</app_version>

</app_info>
13) (Message 2758)
Posted 15 Mar 2014 by Profile [AF>Amis des Lapins] Phil1966
Post:
Haswell can crunch one task in abou 720s with avx version.


Hello !

Do you mean : Faster than the GPU APP ????

Intended to do some tests on AVX this week-end, but as the project is down for maintenance, I took some time to read the forums, and I found the above post ...

Will post the results here, but I cannot believe that 1 task can be crunched in 720s, faster than a CUDA task crunched by a GTX TITAN.

Hereafter the details of the crunchbox :

MB = Asus Z87 Pro
CPU = i7 4770K @ max 4.4 - HT OFF !
WC = Nepton 140XL
GPU = GTX770 Gigabyte 4 Go
RAM = 2 x 8g Corsair Vengeance 1600 etc
W7U (Unfortunately, have not installed Ubuntu on this crunchbox, yet)

Kind Regards,

Philippe
14) (Message 2273)
Posted 30 Dec 2013 by Profile [AF>Amis des Lapins] Phil1966
Post:

Regarding AVX vs SSE3, on an i73770K there seems to be from 3% to about 15% difference in run times. I expect you would need an ix-4xxx CPU to benefit fully from AVX code developments


Yes. AVX on 3xxx stuck on memory loading (CPU does two SSE3 op). AVX is for 4xxx CPUs.

If the WU's for the GPU and the CPU are the same, just different apps, then a GTX770 is twice as fast as a single i7 CPU thread, but only has 1/4 the performance of the entire CPU.


Yes. It is.


Then I don't understand why you launched this GPU app ?
Is it a test for a futur "different / new type of" WU's ?

Have tryed AVX WU's on an i7-4770K ... but I don't know why, very very slow +
only increasing the CPU temp. I guess one should run AVX on 4 cores only => no real gain <-> SSE3 on 8 cores ...

Wish you a Happy New Year !
15) (Message 2253)
Posted 29 Dec 2013 by Profile [AF>Amis des Lapins] Phil1966
Post:
OK Thank You.

I understand now ;)
16) (Message 2246)
Posted 29 Dec 2013 by Profile [AF>Amis des Lapins] Phil1966
Post:
Thank You

However, BOINC can't detect AVX CPU's running under Windows ...
17) (Message 2243)
Posted 29 Dec 2013 by Profile [AF>Amis des Lapins] Phil1966
Post:
My friends, please beware AVX code on Hotwell = increased VID = more heat.
Mine is pulling 1.277V for 4.1GHz and temps shot up +10c (yeah, I'm lazy to optimize... :) ).

Happy crunching!


EDIT: Maybe Radim can possibly offer an option in Preferences page? "Enable AVX computing", perhaps?



Hello

Thank you for the link.

In order to receive only AVX, what should I do ?

Put the SSE2 and SSE3 at 0 ?

Thank You

Kind Regards

Philippe
18) (Message 2242)
Posted 29 Dec 2013 by Profile [AF>Amis des Lapins] Phil1966
Post:
Thank You Very Much !
Philippe
19) (Message 2238)
Posted 29 Dec 2013 by Profile [AF>Amis des Lapins] Phil1966
Post:
Wanted to send you this message via PM ...

Hello,

I have another question : how can I force receiving AVX CPU UT's ?

Thank You

Philippe

NB Will leave the GTX 770 on the project tonight, if this helps ;)
20) (Message 2234)
Posted 29 Dec 2013 by Profile [AF>Amis des Lapins] Phil1966
Post:
Hello !

Thank You

I will reload some for my GTX770 + GTX570M=> 2 more for your stats ;)

Thank You

Kind Regards

Philippe


Next 20