Posts by Crunch3r

1) (Message 5246)
Posted 14 Mar 2017 by Profile Crunch3r
Post:
sorry bout that.

here's another one to try.

http://www.boincunited.org/opt_apps/period_search_10210_x86_64-pc-linux-gnu__avx128_v2.tar.bz2
2) (Message 5236)
Posted 12 Mar 2017 by Profile Crunch3r
Post:
Thanks, but the link does not work, I get a "no hotlink" warning :(

http://www.boincunited.org/images/no_hotlink.gif

Edit: Works with MS IE, no clue why.


https://de.wikipedia.org/wiki/Hotlinking
3) (Message 5226)
Posted 10 Mar 2017 by Profile Crunch3r
Post:
So, if the problem is lack of disk space, how much money does it take to double the disk capacity? This is not a sarcastic question. I'm serious. Is this just a hardware problem or are there software issues as well?


It is just hardware problem. Server has capacity 3.8 TB and if there are many WUs then the free capacity is able to rapidly go down.

And I suppose that it would cost around 5000 € because it would be best to have independly server with higher capacity just for results and all other files which takes so much space. There are also other non-BOINC server processess which could run on the other server. The current server also have no free disk slots so if we just replaced hard drives for bigger than we would have many of them with with no use without another server.


Would you please consider creating a 'Campaign' over at http://www.bitcoinutopia.net/bitcoinutopia/index.php to get some Bitcoins to fund some new hardware ?
4) (Message 5209)
Posted 9 Mar 2017 by Profile Crunch3r
Post:
Hi,

this ones for linux 64 bit.

http://www.boincunited.org/opt_apps/period_search_10210_x86_64-pc-linux-gnu__avx128.tar.bz2

if that one works, i'll compile one for windows too.
5) (Message 4925)
Posted 27 May 2016 by Profile Crunch3r
Post:
wow, this looks very good.. incredible performance, perfect consumption and good pricing..

GTX 660Ti does 2460 SP GFLOPS pulling 150watts max
GTX 770 does 3213 SP GFLOPS pulling 230watts max
GTX 780 does 3977 SP GFLOPS pulling 250watts max
GTX 970 does 3493 SP GFLOPS pulling 145watts max
GTX 980 does 4612 SP GFLOPS pulling 165watts max
GTX 1070 will do 6500 GFLOPS SP pulling 150 watts and MSRP of $379.
GTX 1080 will do 8227 GFLOPS SP pulling 180 watts and MSRP of $599.

You can replace a 660Ti, 770 and 780 with 2x GTX 1070 and gain 34% more GFLOPS and use less than 50% of the power. :)


Those rating are single precision (FP32) and are useless here, since Asteroids uses double precision(FP64)...

AFAIK, a gtx 1080s FP64 performance is 1/32 FP32 peak flops...
A GTX 660TIs FP64 performance is 1/24 FP32...
6) (Message 3973)
Posted 24 Jan 2015 by Profile Crunch3r
Post:
Hi Crunch3r,

Obliged to you for the assistance, got the app and its working ok:-)

Regards, Cliff


Nice,

I'm glad I could help.
7) (Message 3954)
Posted 18 Jan 2015 by Profile Crunch3r
Post:
well, the easiest way to deal with it, is downloading that app -> http://www.boincunited.org/period_search_10210_windows_x86_64_bd_fma4_gcc.zip
using an app_info.xml. No doubt about it, it's the fastest app you can run on your cpu. since it's using fma4.
8) (Message 3045)
Posted 16 May 2014 by Profile Crunch3r
Post:
OK, to anyone who want's to try the BD fma4 app

here is the link.

http://www.boincunited.org/period_search_10210_windows_x86_64_bd_fma4_gcc.zip

It's using anonymous platform and the only thing to do is to copy it to the project directory.

I won't go into specifics on how to install the app, since only experienced boinc users should have a go at it.
9) (Message 2861)
Posted 21 Mar 2014 by Profile Crunch3r
Post:
Is there any speedup over avx?


AVX was always slower than the sse2 or sse3 app. Not even using the intel compiler could fix that.

So i've switched to gcc 4.8.2 and recoded the parts of the app that could make use of amds fma4.

The real question is, if the app is faster than the sse2 one (fastest of them all), however , as i've mentioned earlier, my cooling sucks and my fx 6100 keeps trottling so there's no way of me telling what the differece is.
10) (Message 2829)
Posted 18 Mar 2014 by Profile Crunch3r
Post:
Is someone interested in experimenting with a app that uses the AMD fma4 instruction set ?

FWIW, i've been running it for a couple of weeks now at it works (valid results).

If someone is interested i could upload it.

Let me know.


Did you recode the app or change compiler settings only? I have FMA2 app and not difference between AVX and FMA2.


I rewrote the code.
11) (Message 2813)
Posted 17 Mar 2014 by Profile Crunch3r
Post:


AAAH - noob alert! ;)

currently i got no AMD-FX at hand, but would you throw in some performance-numbers?


I wish i could throw in some numbers but unfortuantely, my cooling is not efficient enought.. (meaning my fx6100 keeps throttling down to 1.4 GHz)... So it's a guessing game.

So, the only available option is testing it on some bullozers that can handle the heat ;)
12) (Message 2809)
Posted 17 Mar 2014 by Profile Crunch3r
Post:
Is someone interested in experimenting with a app that uses the AMD fma4 instruction set ?

FWIW, i've been running it for a couple of weeks now at it works (valid results).

If someone is interested i could upload it.

Let me know.
13) (Message 2770)
Posted 15 Mar 2014 by Profile Crunch3r
Post:
And we don't yet know if the person/persons involved have 24/7 connectivity.


http://www.planetaryresources.com/

I'm pretty sure they are online 24/7...


I figure those 130 machines must be servers therefore assume 24/7 connectivity.


You're kidding me ?!? Of course those machines are servers !


If he has permission to install BOINC on all of them then I doubt his account's connectivity is restricted.


I'm guessing it's in their best interest to run asteroids@home given the link i posted above... Guess what that company is interested in ...


Even a restriction of 20 cached tasks per core would be a big help in solving the problem. If one assumes ~1 hour per task then 20 tasks per core should be enough to last from 1 connected window to another. It wouldn't be perfect for everybody but the pros definitely outweigh the cons. Remember back to several months ago when Kyong made a mistake on the task duration estimate and everybody got 10 times as many tasks as they should have. The kind of restriction we're talking about would have helped a lot in that situation too.


Bla bla bla... just limit the "in progress WUs" per core to let's say 15... that'll solve the "problem" for those that can't wait... Instant credit would't that be cool ?!?... Not going to happen here... So why don't you all chill till you get your credit... ?!?
14) (Message 2551)
Posted 18 Feb 2014 by Profile Crunch3r
Post:
My answer from PM, just corrected some typos:

Well, we'll see, what reports will be at forum. Without the hardware it is difficult to check it even just looking through source codes. I was even ready to disable all mac os apps because there was a new problem. We have many bad results which were computed by os x apps, but there is a problem which validator couldn't verify. Now it is fixed, so there are more invalid results from os x than before. But not all. For compilation was used the same source code which is for linux and win, so I don't know where is the problem and I can't find it out. I will try to check what will be reported at forum but solving os x problems have the minimum priority for now. We also wrote some e-mails to Apple for some hardware for scientific purposes but got no answers so they probably don't care about it if we got no responses from them. And buying appropriate hardware isn't cheap, so the os x apps are now with minimum priority.


The problem is the OSX avx app (Intel 64-bit Mac OS 10.5 or later 102.10(avx_osx) ) ,which was compiled using llvm. I've send you a PM couple of weeks ago,that it needs to be removed! Somehow that didn't happen. I do have a new OSX avx app here that validated on my xeon 1240, however it's way slower than the sseX OSX app, so no need for a avx app on osx unless you wanna waste time!
15) (Message 2058)
Posted 18 Nov 2013 by Profile Crunch3r
Post:
Thanks for pointing on 780ti dp speed. I thought it's like titan and it is not.

We need cuda 5.0 and better because of linker. I think cuda version do not affect speed on older cards. It's more related to code than to cuda version.


Ok. So let's be honest... If we're porting asteroids to gpu, the only way to go is opencl! That will of course take into account that all nvidia gpus except the titan suck at DP...

So no need to waste energy in porting the app to cuda when we already know that asteroids will be dominated by ATI/AMD gpus because those are less crippled regarding DP performance.

To face the fact, Asteroids will be a "second" Milkyway@home regarding GPU dominance.

AMD/ATI GPUs will dominate and nvidia will be left biting the dust.
16) (Message 2028)
Posted 7 Nov 2013 by Profile Crunch3r
Post:
I have released the optimized versions for OS X, which were compiled by Crunch3r.


I've compiled new apps 32 and 64 bit.
There seems to have been an issue with the sse2 app getting invalid results.

I'm got a hunch that it's due to using clang instead of g++.
anyway, i've uploaded the new apps(see below).

http://www.boincunited.org/opt_apps/asteroidsathome_osx_10211_32bit.tar.bz2
http://www.boincunited.org/opt_apps/asteroidsathome_osx_10211_64bit.tar.bz2

(no need for sse2)
The avx app might follow later since i have to figure out why it's so slow compared to the sse3 app.



17) (Message 1982)
Posted 26 Oct 2013 by Profile Crunch3r
Post:
OK.. i've compiled sse2, sse3 and avx apps. For the moment OSX 10.5 &64 bit only.

32 bit apps will follow later.

Now it's up to Kyong to make them available ;)

regards
Crunch3r


the 32bit app is uploaded as well. It's 'only' sse3 since avx is way slower.
18) (Message 1981)
Posted 26 Oct 2013 by Profile Crunch3r
Post:
OK.. i've compiled sse2, sse3 and avx apps. For the moment OSX 10.5 &64 bit only.

32 bit apps will follow later.

Now it's up to Kyong to make them available ;)

regards
Crunch3r
19) (Message 1973)
Posted 25 Oct 2013 by Profile Crunch3r
Post:
Great, thanks. Yes, both. And what about SSE2? I have no view of current CPUs on Mac.


No need for SSE2 since all macs have at least sse3 available.
20) (Message 1971)
Posted 25 Oct 2013 by Profile Crunch3r
Post:
I'll compile a new binary for OSX.
It'll be available for download tomorrow.

Do we need both sse3 and avx ?


Next 20