Posts by Raistmer


log in
1) Message boards : Number crunching : Seti fundraising via Bitcoin Utopia (Message 1603968)
Posted 1 day ago by Profile Raistmer
The SETI@home Campaign is going live this weekend.
http://www.bitcoinutopia.net/bitcoinutopia/forum_thread.php?id=622#6491

Where they post rules for such events? What exactly participant should do to bring money to SETI@home? Would be nice to see such guidance for absolute novice too.

EDIT: and what control mechanisms to check if donated computation time was converted to money for SETI@home as expected ?
2) Message boards : News : Running out of workunits (Message 1603462)
Posted 2 days ago by Profile Raistmer
Just an idea but have you all ever considered trying one of the crowd funding site to get donations for the sole purpose of buying time on telescopes for the Seti project.
If people will pony up 5 buck to help people build a "better ice chest" to the tune of a million plus, I am sure there would be tremendous interest in funding
Seti research.
Plus you would be reaching a whole new group of people(young) who you have never reached by your other methods.
Then there is the "geek" factor,seems that these days people like to help geeks.


Worth to consider indeed.
3) Message boards : Technical News : What's happening... (Nov 8, 2014) (Message 1603168)
Posted 3 days ago by Profile Raistmer
Да уж, пичалька:(,а я только стал SETI считать

Два года после регистрации раскачивался что ли, с 2012? :)
Ничего, скоро дело пойдет!
4) Message boards : Technical News : What's happening... (Nov 8, 2014) (Message 1602671)
Posted 4 days ago by Profile Raistmer

Добрый день. Ответьте, пожалуйста, кто знает. :-) С одной стороны вот тут говорят, что только астропульса нет, но у меня уже третий день вообще никаких заданий не приходит. В логах пишется, что типа нет заданий для скачивания. Это у меня что-то не так, или уже _все_ задания могли кончиться? Спасибо.

В данный момент вообще ничего нет, к сожалению. Скоро должна раздача модулей возобновиться. "Все" кончится еще не скоро, сейчас просто проблемы с серверами.
5) Message boards : Number crunching : Best performing hardware (Message 1602518)
Posted 4 days ago by Profile Raistmer
[quote]
Then there's the memory problem -- our particular model only has 8 GB of ram, which has to contain the OS as well as applications. At the moment top reports 7.6 GB free, so for 60 cores x 4 threads that'd be only around 30 MB available per thread; currently on an Ubuntu box s@h reports 104 or 164 MB virtual memory per process, 40 or 96 MB resident per process, and 12 MB KB shared. (Similar figures on a RHEL box; the one job on Ubuntu that's taking more RAM is not the vlar WU running at the moment.)
This might all change with the new Phis, which slot into motherboard sockets and can access main memory, but Intel hasn't offered me one to play with yet...

Would it be possible to use massively-parallel system via OpenCL drivers and not to run separate instance on each node/core ?
6) Message boards : Number crunching : Seti fundraising via Bitcoin Utopia (Message 1602517)
Posted 4 days ago by Profile Raistmer
Just out of curiosity, how can you tell if the value of the coins you mine is greater than the cost of the electricity you use to mine them?


Some can't or don't want to donate money directly but gladly donate electricity via computations. If project need money more that raw computations and some another project can give money in exchange of computation then that look good bargain even if energy spent will cost more than project in question (SETI in out case) will recive as real money.

Of course would be better to donate directly but... good to have some other options too. Also, some of BOINC participants bought specialized BitCoin ASIC's.
That hardware can't be used in SETI directly in any way. But via such fundrising even this piece of useless silicon (at least from SETI-fan point of view) can be used ;)
7) Message boards : Number crunching : Time to warm up back up projects? (Message 1601233)
Posted 9 days ago by Profile Raistmer
Time to extensive offline benching and tuning. Also system cleanup and reconfiguration. All that maintenance can take some time ;)
8) Message boards : Technical News : What's happening... (Nov 8, 2014) (Message 1601221)
Posted 9 days ago by Profile Raistmer
Так, насколько я понял, какое-то время данных для обработки не будет, правильно?
Есть ли примерные сроки возобновления работ?

Пока есть что считать. АстроПульса только нет. МультиБим скоро может закончиться.
Как починят, так и продолжим. Сервак накрылся и погубил базу данных в процессе...
9) Message boards : Number crunching : Comp Errors with New Intel Drivers - Beware (Message 1600791)
Posted 10 days ago by Profile Raistmer
Would be good if such findings would be reported also back to Intel via their forums or customer support. This would allow at least some hope that next driver update will be not so disastrous.
10) Message boards : Technical News : Taking longer than expected (Nov 13, 2014) (Message 1600790)
Posted 10 days ago by Profile Raistmer
If complete rebuild inevitable maybe it's good time to think about different database engine, more optimal one (if such exits, of course...)
11) Message boards : Number crunching : Intel and Nvidia run together?? (Message 1600204)
Posted 11 days ago by Profile Raistmer
http://setiathome.berkeley.edu/show_host_detail.php?hostid=6915997

IvyBridge, iGPU enabled and used, GSO9600 + GT9400 both enabled and used for SETI.
So, yep, it works.
12) Message boards : Technical News : What's happening... (Nov 8, 2014) (Message 1599829)
Posted 12 days ago by Profile Raistmer
You talk about the "increasing the minimum quorum" and "keep all of the results going to the same kind of hardware" effect.

But what about "a mechanism that sends out overflow results for 2nd round of crunching"?
I understand this as:
If validator have 2 "overflow results" ('initial replication') and they match - do not mark the WU as valid - instead send 3-rd task to check is that real or false overflow


Yeah, would be good addition! And here "to send it on different hardware type" will be most appropriate.
Unfortunately, to implement such logic definitely additional coding on validator side required. Will see if someone could do this in near time but idea is quite good IMHO, worth to elevate it for review to SETI core team.
13) Message boards : Technical News : What's happening... (Nov 8, 2014) (Message 1599636)
Posted 13 days ago by Profile Raistmer
I suppose increasing the minimum quorum could be used if they wanted to have more results with the same data. Depending on how closely the apps look at the data chunks. I could see requiring 3 or 4 results as still be scientifically legitimate, & not considered "busy work".

Especially to combat those situations where two GPUs get paired up together and spew out -9 overflows when it shouldn't have been a -9 overflow WU. I know that situation isn't frequent, but it also isn't uncommon/rare.

I think to fix that extra code would have to be added to keep all of the results going to the same kind of hardware. I don't think having 3 or 4 GPUs spit out a false -9 overflow would be very helpful.
Perhaps a mechanism that sends out overflow results for 2nd round of crunching or something along those lines.
t

yep,it will prevent such type of errors. But will increase inconclusives rate ultimately. Same hardware tends to round results in the same way. But that way is different between types of hardware. Hence, with your proposal, more results rounded differently will be paired agains each other. And it will increase rate of inconclusives with additional task issued to resolve inconclusive.
There are always some signals too close to threshold to feel even slightest difference in rounding.
14) Message boards : Number crunching : why no graphics/screensaver for v7 7.03 (opencl_intel_gpu_sah) files? (Message 1599580)
Posted 13 days ago by Profile Raistmer

Something comes to mind by the way, I see that the OP is specifically asking about the graphics/screen saver on an Intel GPU. These do tend to lose the bottom part of the graphics, because of what appears to be a problem with the OpenGL module in the drivers that Intel have released. Perhaps that that was his question and we've now totally gone off topic without answering him on that.


You missed this word in thread title: opencl_intel_gpu_sah
That means OpenCL iGPU MultiBeam app in question and we totally on topic explaining why graphics not implemented for such type of applications in whole and for this particular as example.
15) Message boards : Number crunching : why no graphics/screensaver for v7 7.03 (opencl_intel_gpu_sah) files? (Message 1599565)
Posted 13 days ago by Profile Raistmer

I can view graphics from an OpenCL application running on an Intel GPU (matching the question which opened this thread), from a CUDA application running on an NVidia card, and I can view the iGPU graphics on a monitor connected to an NVidia graphics card - so it's all device-independent.

1)Does it show something depending on current dataset in progress?
2)What performance impact?
16) Message boards : Number crunching : why no graphics/screensaver for v7 7.03 (opencl_intel_gpu_sah) files? (Message 1599553)
Posted 13 days ago by Profile Raistmer
OpenCL - OpenGL interoperability may come into play, but it's separate field to study, not connected with computations at all.

OpenCL uses the same hardware device components as OpenGL, it's just a different language for programming.


It was not about hardware, it's about possibility to pass GPU memory buffer handle from OpenCL to OpenGL... running in another process adress space.
17) Message boards : Number crunching : why no graphics/screensaver for v7 7.03 (opencl_intel_gpu_sah) files? (Message 1599397)
Posted 13 days ago by Profile Raistmer
As Richard already stated to show something in screen saver that something should be copied into separate process address space.

1) I'm not aware about easy ways to share GPU device memory buffer between different processes. OpenCL - OpenGL interoperability may come into play, but it's separate field to study, not connected with computations at all. And definitely such interoperability (even if possible between 2 separate processes at all) has nothing to do with described (by cited link) on BOINC Graphics API page. Hence, as already was absolutely correctly stated, efficiently enough way to add graphics screen saver incorporates additional development efforts completely disconnected with main task.

2) BOINC Graphics API assumes data to show in system memory. Hence one need to bring that data to system memory (of another process need to add). To do so one needs to just copy corresponding data buffer (and form it first from real data buffers used in app) from one address space to another. Or, even not copy but just map it (system shared memory). Mapping is relatively cheap operation so main processing load on CPU is to get frame buffer from real data buffer (non-scientific load on CPU, can be big or not for particular apps).
That's how things are in CPU-based apps.
For GPU app the single way (on modern non-APU or non-HSA, discrete GPUs) to compute efficiently is to keep all data in device memory, not in system memory.
Hence, to present some data to screen saver one needs to bring some data from GPU device memory back to system memory, then to form frame buffer to show, then to pass/map it to another process address space and only then to use BOINC API to "simply show" buffer inside graphics app process( maybe BOINC API includes mapping too but this doesn't change the picture). Link between system memory and GPU is highly assymetrical. Even CPU-> GPU path slow enough comparing with CPU-CPU memory transfers. But GPU->CPU one even slower. Hence, to show something really interesting connected with calculations done by app and not just static picture inside screen saver one should considerably hinder real processing (with "easy way to accomplish") or put additional efforts in developing adequate memory transfers overlapping with real computations, that again constitutes separate development effort not in slightest degree covered by provided BOINC Graphics API link.

Interesting in implementation - welcome to team. As was noted, sources are openly available.
18) Message boards : Technical News : What's happening... (Nov 8, 2014) (Message 1599013)
Posted 14 days ago by Profile Raistmer

4. Old Data. We have been working on old data. It's not a "make work" thing. Most of the data we've been sending had a problem the first time around. Either part of the data was left unprocessed, or the results were questionable. In addition, all the old data that has been sent had not been processed with S@H v7, so there was no autocorrelation analysis done on it.


Is it possible or worthly to turn off gaussian, triplet and pulse finding on those tapes for re-processing that had no other issues than missing autocorr at that moment?
Or smth another like chirp rate and blanking algorithm change makes all types of search worthy even on that subset of tapes?
19) Message boards : Number crunching : BOINC not always reports faster GPU device... (Message 1599008)
Posted 14 days ago by Profile Raistmer
* Neither likes to be criticised in public

QED. Point made. Next time you want David to change something, post where he will read it - and put yourself in his shoes, before you write it.

I don't want, I'm fine with this issue as it is cause use corresponding flg in cc_config.xml. But I think issue worth to be recorded in case someone stumble on it and will be trapped. What I want to be fixed I write in BOINC dev list. With almost same outcome as to write on house wall in Moscow :P
20) Message boards : Number crunching : BOINC not always reports faster GPU device... (Message 1598997)
Posted 14 days ago by Profile Raistmer
I'm afraid response was provoked indeed, but such that would not lead to any useful discussion. At least from my side. Issue described. Who interested may continue.


Next 20

Copyright © 2014 University of California