Posts by Raistmer


log in
1) Message boards : Technical News : Latest Astropulse Status (Message 1649125)
Posted 1 hour ago by Profile Raistmer
Can we get status update, please?
2) Message boards : Number crunching : Intel® iGPU AP bench test run (Message 1649066)
Posted 5 hours ago by Profile Raistmer
...
Joe, I guess you meant 1520/760 instead of 1520/750.

Yes, fingers and mind lost sync.
2nd run of 1472/736 and ~ 3 secs less elapsed time, but ~ 8 secs more CPU time. Didn't thought that a so big difference would be possible between two runs.

That's the kind of variation which Jason's suggestion would characterize. But 30 runs of even the best 4 pairs of ffa settings seen so far would be a day and a half of steady testing, in my view not justified yet.

I suggest the next step check different ratios between -ffa_block and -ffa_block_fetch, specifically 736/736, 2208/736, 2944/736, 1472/1472, 1473/491, and 1472/368.
Joe


Keep an eye on counters values from stderr. Sharp increase of "misses" would mean some issues with particular params set.

EDIT: bolded value can lead to driver restarts.
There is no sense to find optimum in odd values taking into account that wave for iGPU is even number. Though in most kernels there is 2D launch domain some of them use 1D domain (so directly misconfigured number of waves) and some use odd secondary dimension (again, misconfigured domain size in case of odd first dim size).

EDIT2: does this task contain any rep pulses ? If yes, this fine tuning is void. Use Clean* tasks instead. As I said penalty from single miss is big enough. Current design tries to pre-compute whole ffa_block number of periods. If single period have signal all periods will be firstly re-processed on GPU then part of those periods will be shortly viewed by CPU also.
3) Message boards : SETI@home Science : Strange Question: Time Difference between stars (Message 1649049)
Posted 6 hours ago by Profile Raistmer

If you travel above speed of light

Too big "if" to reside in modern SETI@home science section. For now it belongs to "wishes and dreams" section mostly.
4) Message boards : Number crunching : The Highest Ranked SETI AMD Host is a MAC: Time for a STOCK MAC APP? (Message 1648890)
Posted 16 hours ago by Profile Raistmer
Yep. And all those changes in sources themselves that consistute new revision :)
5) Message boards : Number crunching : @Pre-FERMI nVidia GPU users: Important warning (Message 1648885)
Posted 16 hours ago by Profile Raistmer
Any confirmation that new build works OK with 341.44 driver on pre-FERMI cards?

Would you consider this a release candidate? If so, I'll dig the ancient 9800GT out of storage yet again and run the bench. If we're not near a release, I'll leave it a bit later to test any other tinkerings as well - I have to re-rig the supplementary power supply too, so it's all a bit of a hassle.


There were no updates in AP tree since last beta. So yes, if this app works Ok it will be supplied to Eric for upload on beta and to you to include in installer update.
6) Message boards : Number crunching : @Pre-FERMI nVidia GPU users: Important warning (Message 1648520)
Posted 1 day ago by Profile Raistmer
Any confirmation that new build works OK with 341.44 driver on pre-FERMI cards?
7) Message boards : SETI@home Science : Strange Question: Time Difference between stars (Message 1648479)
Posted 2 days ago by Profile Raistmer
I am trying to understand what the time difference between starts and their respective plants.

For example, Lets assume that right now it's 6:00 PM "Earth Time" throughout the entire planet. In fact lets say there is only 1 time zone for all of earth.

What time would it be on mars?
What time would it be on Jupiter?
What time would it be on Pluto? (yes I know it got fired)


Now assuming my question wasn't entirely stupid, what time would it be:

On Sirius?
On any planets who "orbit" Sirius?


I am trying to figure out what the change in time is relative to the start itself.

Thanks

Robert


I would recommend to read corresponding chapter of this book: Roger Penrose. «The Emperor's New Mind»

It describes quite visually how time depends on velocity of observer. For galaxy-scale ranges (that is, event occurs in another galaxy from us) "simultaneity" will be different even for 2 mans walking towards each other.
In any case the sense of simultaneity is very abstract for пространственно-подобный interval.

EDIT: Even Wiki page illustrates it well enough: https://en.wikipedia.org/wiki/Relativity_of_simultaneity

EDIT2: Actually, Wiki has all needed parts:
people pass each other on the street; and according to one of the two people, an Andromedean space fleet has already set off on its journey, while to the other, the decision as to whether or not the journey will actually take place has not yet been made. How can there still be some uncertainty as to the outcome of that decision? If to either person the decision has already been made, then surely there cannot be any uncertainty. The launching of the space fleet is an inevitability. In fact neither of the people can yet know of the launching of the space fleet. They can know only later, when telescopic observations from earth reveal that the fleet is indeed on its way. Then they can hark back to that chance encounter, and come to the conclusion that at that time, according to one of them, the decision lay in the uncertain future, while to the other, it lay in the certain past. Was there then any uncertainty about that future? Or was the future of both people already "fixed"?

—Roger Penrose, The Emperor's New Mind: Concerning Computers, Minds, and the Laws of Physics

https://en.wikipedia.org/wiki/Rietdijk%E2%80%93Putnam_argument
But I would recommend to read the original book though, there is lot interesting, much more than just this single time relativity illustration.
8) Message boards : Number crunching : Intel® iGPU AP bench test run (Message 1648477)
Posted 2 days ago by Profile Raistmer

The whole bench test run could lasts days, I want to find the best params. ;-)

One need understand that best params set depends from data set in particular task.
App computation flow consists of compromises of type "do this faster usually but slower if rare event occurs".Rare event is signal found (and best signal update in case of MultiBeam).
Hence to see true best option one needs to collect great statistics with different blanking areas, different number of reported pulses and so on.
In short, hardly possible for offline runs.
This means at some point small differences that show in artifical test on silenced task will not cover situation with real workunit.
For example, one could found that bigger ffa_block sizes take less time on silenced task. But in case of signal found in such big chunk of data the time penalty for re-processing this big chunk will kill all benefits accumulated for whole duration of task run. While smaller chunks will locate origin of signal more precise and allow less penalty on reprocessing.
9) Message boards : Number crunching : The Highest Ranked SETI AMD Host is a MAC: Time for a STOCK MAC APP? (Message 1648289)
Posted 2 days ago by Profile Raistmer
You have binary compiled by others. It should accept some params described in ReadMe. Did you try to tune that params (namely, period_iterations_num) to reduce lags?
10) Message boards : Number crunching : The Highest Ranked SETI AMD Host is a MAC: Time for a STOCK MAC APP? (Message 1648254)
Posted 2 days ago by Profile Raistmer
If an App makes your typing stutter by a second or two, it probably won't get used much.

Did you try all builtin tuning capabilities before building own binary?
11) Message boards : Number crunching : The Highest Ranked SETI AMD Host is a MAC: Time for a STOCK MAC APP? (Message 1648131)
Posted 3 days ago by Profile Raistmer
when I said prev rev I meant whole rev sources, not just single CL file.
HD5 triplet was added in rev 2799 so use 2763 to avoid them.
r2850 and r2849 had some changes in OS X build path though, maybe worth to merge that changes too.
12) Message boards : Number crunching : The Highest Ranked SETI AMD Host is a MAC: Time for a STOCK MAC APP? (Message 1647735)
Posted 4 days ago by Profile Raistmer
It appears the "end FindSpikes_ev==NULL" comes and goes. Anyone know where they can be coming from?
...

Build features: SETI7 Non-graphics OpenCL USE_OPENCL_HD5xxx OCL_CHIRP3 ASYNC_SPIKE FFTW SSSE3 64bit
System: Darwin x86_64 Kernel: 14.1.0
CPU : Intel(R) Xeon(R) CPU E5462 @ 2.80GHz

Try to disable ASYNC_SPIKE define.
This path not fully mainained now cause I don't use it in last Windows builds. "should" work though...

EDIT: Also, better to experiment with prev revs, w/o PC_find_triplets_*_HD5_cl kernels. They still under development and known to work incorrectly in last build.
13) Message boards : SETI@home Staff Blog : Sending Messages to ET: Just say no, for now. (Message 1646897)
Posted 5 days ago by Profile Raistmer
-Governments will hide the fact a signal or responce had been made in a large attempt to stop faiths going out of control.


I still don't understand why people think this stuff. Governments can't cooperate on much of anything, yet people think they will agree to hide an ETI signal from the entire world. I just don't see it. The bigger the secret, the harder it will be to keep. Someone in the gov't will speak (to anyone - a family member; a friend; a confidante), and the word will get out, and evidence will be found after enough prying and digging.


Don't make me laugh. What evidence? There are whole seas of "evidencies", just look in UFO area. But you refer solid evidence, perhaps. Not rumors like will be if someone say something to someone else. But look, peoples can't get right picture what really going on in another countires yet you claim there can be "some evidence" found of Contact... Rumors all one could hope to get... [in case if decision to hide info will be committed, of course]
14) Message boards : Number crunching : @Pre-FERMI nVidia GPU users: Important warning (Message 1646894)
Posted 5 days ago by Profile Raistmer
fine-grained driver version checking added: https://www.dropbox.com/s/v8g8a4la4j6osk5/AP7_win_x86_SSE2_OpenCL_NV_r2745.7z?dl=0
15) Message boards : Number crunching : The Highest Ranked SETI AMD Host is a MAC: Time for a STOCK MAC APP? (Message 1646883)
Posted 6 days ago by Profile Raistmer

I still don't understand why placing my 3.3.4 in the OSX64 folder didn't work...it should have.

Perhaps because we use statically linked FFTW most of time. That means you need to link object code with proper library on link stage instead of trying to make already build binary to load some new shared file.
In Windows we have .dll for shared objects and .lib for libraries (that contain object code for linking).
In *nix systems there are .so and .a AFAIK. So, you need to provide right .a file on build stage. That's what Claggy and Jason suggested you to do. Putting .a file into system directory and hoping that binary will pick it up after linkage completed is not a way to go.
16) Message boards : Number crunching : @Pre-FERMI nVidia GPU users: Important warning (Message 1646331)
Posted 7 days ago by Profile Raistmer
Raistmer:
Can you please do some testing on your end, using 341.44, to verify it's working for you?

Build with lifted ban for 341.xx drivers: https://www.dropbox.com/s/uedoe9qoxnqy6qu/AP7_win_x86_SSE2_OpenCL_NV_r2745.7z?dl=0
17) Message boards : Number crunching : The Highest Ranked SETI AMD Host is a MAC: Time for a STOCK MAC APP? (Message 1646281)
Posted 7 days ago by Profile Raistmer
It works!

So, it won't build in Lion, builds but doesn't run in Mountain Lion, Runs in Yosemite. Well, it's always something.

Congrats!
Apple little rot it seems :)
18) Message boards : Number crunching : The Highest Ranked SETI AMD Host is a MAC: Time for a STOCK MAC APP? (Message 1646072)
Posted 7 days ago by Profile Raistmer

BTW, all those other files such as;
seti_boinc-CLInfo.o seti_boinc-GPU_lock.o seti_boinc-analyzeFuncs.o seti_boinc-analyzePoT.o seti_boinc-analyzeReport.o seti_boinc-autocorr.o seti_boinc-chirpfft.o seti_boinc-fft_execute.o seti_boinc-fft_kernelstring.o seti_boinc-fft_setup.o
Are in the AKv8/client folder.


Похоже, скрипт просто испорчен. seti_boinc - это каталог. А у тебя это часть имени файла. Проверяй расстановку кавычек...
19) Message boards : Number crunching : The Highest Ranked SETI AMD Host is a MAC: Time for a STOCK MAC APP? (Message 1645998)
Posted 8 days ago by Profile Raistmer

It would be nice to understand why I'm getting all these seti_boinc no such file statements.

Cause that directory mentioned in scripts you use.
Why it still mentioned there is another question. PErhaps, because used script old and unmaintained.
Maybe Urs could point to more fresh build scripts to use.
20) Message boards : Number crunching : The Highest Ranked SETI AMD Host is a MAC: Time for a STOCK MAC APP? (Message 1645982)
Posted 8 days ago by Profile Raistmer

Well, every MB .cl file I've seen has numbers, including the ones posted by Joe Fox compiled on a Mac in the Terminal, MultiBeam_Kernels_r2760.cl. Are you saying people are just adding those numbers onto the file name?


Yep, they do. [BTW, petri33 pointed that twice already] Unless they devise script to do that from SVN rev number.
It's done for Windows (look corresponding .bat files) and probably for Linux too. If you have some SVN command line tools look to mentioned .bat files and SVN tools manual to devise something similar for for OS X. Script should look into SVN revision number of sources then copy MultiBeam_Kernels.cl file into same directory where executable created after building process and name it MultiBeam_Kernels_rN.cl where N is revision number from prev step.

It can be done manually of course but it's boring so script preferable.

Wouldn't be useful to attempt to get actually working configure and build scripts from Joe Fox who seems had compiled to working binaries already.

AFAIK Urs has launchable binary for OS X too but there is precision issue currently.


Next 20

Copyright © 2015 University of California