Posts by petri33

1) Message boards : Number crunching : GTX is dead long live RTX !! (Message 1953899)
Posted 16 days ago by Profile petri33
Post:
Hi Petri, will you be able to "virtual" test on a Turing card without physically having one in your possession? Or will you have to debug on actual owners cards when they attempt to use the special app?


I could use X-term window remotely if the machine had CUDA development environment and gnu C++ and I could deliver source code and off-line benchmark tool with WUs.
2) Message boards : Number crunching : GTX is dead long live RTX !! (Message 1953877)
Posted 16 days ago by Profile petri33
Post:
Petri has said he is happy with his Titan V and is not looking to upgrade to Turing.
Yes, but those 3 - 1080 series cards are getting worn out by now :D


Two 1080's on the shelf (spare) and two more plus one 1080Ti installed in the machine.
I'd be happy to toss all 1080's and Ti out if could get more Volta at an affordable price.

I may try a TuringTi next summer.

When my old cards fans have worn out and died I'll replace the cards.
3) Message boards : Number crunching : Arecibo Overflow Tasks Run With CUDA Special App Sometimes Filled With Triplets Only (Message 1953875)
Posted 16 days ago by Profile petri33
Post:
This is interesting. I found two BLC Overflows also filled with Triplets;
http://setiathome.berkeley.edu/workunit.php?wuid=3123251218
Download http://boinc2.ssl.berkeley.edu/sah/download_fanout/35a/blc11_2bit_guppi_58227_21475_HIP66640_0059.30937.0.22.45.81.vlar
http://setiathome.berkeley.edu/workunit.php?wuid=3123251224
Download http://boinc2.ssl.berkeley.edu/sah/download_fanout/4c/blc11_2bit_guppi_58227_21475_HIP66640_0059.30937.0.22.45.84.vlar
However, on these the other Apps mostly agree. So, why don't they agree on the Arecibo files when they agree on the BLCs?


I downloaded the first one. I'll run a test. The second one was gone before I could DL it.


The reference on CPU had 23 triplets and then 7 pulses. GPU reported 7 more triplets and stopped at 30/30 before reporting those 7 pulses.

On a noise bomb you can find all kinds of spikes, pulses and triplets. Triplets get more common when more high spikes are found because the odds that three spikes are evenly spaced increases. And the longer the sample the more there is those noise induced spikes. No alien signal, just cell phone, radar, satellite, ... electromagnetic noise from Tellus and some wasted seconds of processing time.
4) Message boards : Number crunching : Arecibo Overflow Tasks Run With CUDA Special App Sometimes Filled With Triplets Only (Message 1953849)
Posted 17 days ago by Profile petri33
Post:
This is interesting. I found two BLC Overflows also filled with Triplets;
http://setiathome.berkeley.edu/workunit.php?wuid=3123251218
Download http://boinc2.ssl.berkeley.edu/sah/download_fanout/35a/blc11_2bit_guppi_58227_21475_HIP66640_0059.30937.0.22.45.81.vlar
http://setiathome.berkeley.edu/workunit.php?wuid=3123251224
Download http://boinc2.ssl.berkeley.edu/sah/download_fanout/4c/blc11_2bit_guppi_58227_21475_HIP66640_0059.30937.0.22.45.84.vlar
However, on these the other Apps mostly agree. So, why don't they agree on the Arecibo files when they agree on the BLCs?


I downloaded the first one. I'll run a test. The second one was gone before I could DL it.
5) Message boards : Number crunching : Arecibo Overflow Tasks Run With CUDA Special App Sometimes Filled With Triplets Only (Message 1953349)
Posted 20 days ago by Profile petri33
Post:
Boincmgr could reschedule a noise bomb (any 30/30 task) to CPU. I think that would be waste of resources.

The definition of a pulse or a triplet has not changed. The special app finds valid pulses and valid triplets that are there in the WU.

There are just so many of them. 30 or more to choose from in a noise bomb. To begin with the data in a noise bomb is invalid. You can not make a valid report from invalid data. 30 anything means the WU has noise and it should be discarded and the computation stopped. Noise bombs have no usable data. They need not report anything. Just label the WU invalid.

Petri
6) Message boards : Number crunching : Arecibo Overflow Tasks Run With CUDA Special App Sometimes Filled With Triplets Only (Message 1953303)
Posted 20 days ago by Profile petri33
Post:
// WAS 11 in V0.97 before b2
#define MAX_TRIPLETS_ABOVE_THRESHOLD 13
that is in pulsefind cuda code.

Someone a long time before me has decided a hard limit. 11 or 13 or some value is the maximum of number of ...
If I make that number bigger it slows down triplet finding but reports more of them. If I make that smaller it misses triplets.

Easy to fix. Make it 11. Miss the rest of them.

The problem doesn't appear to be finding the signals, it's about their classification.

GPU application Triplet count: 30
CPU application Pulse count: 30
The CPU is the reference, so the GPU needs to match it's classification of Pulses, not Triplets.


No It Does Not.

EDIT: There are 30 or more of both of them. END EDIT

There are thousands of signals in a 30/30 noise bomb, When the limit is reached the search stops. It is a noise bomb.
Some software stop at some point and some other software stop at some other point.
Crippling a parallel find requiring serial order is something that no one wants.
We're going to the future.
The after processing software or a new type of WU is needed to check that a noise bomb is a noise bomb which it is anyway.
We need no consensus of by what signals (autocorr, triplets, gaussians, pulses, spikes) or whatever is searched for.
It does not matter if any SW says it is a noise bomb and any of them agree/disagree.
It is a noise bomb.
Let the system send it to a GPU to make someone happy <3.
7) Message boards : Number crunching : Setting up Linux to crunch CUDA90 and above for Windows users (Message 1953300)
Posted 20 days ago by Profile petri33
Post:
"bad arg: -pfl
bad arg: 512"

The flag -pfl N is not supported any more.
Please do not try it with UPPERCASE. It is still there so that I can test it on offline benchmarks and bug hunting.

Same applies to -pfe.

Petri

Hey Petri,
With all these Overflowing BLC11 tasks is there something different we can do to find out why some overflows end up Invalid filled with Triplets?

If anyone can find an Overflow filled with Triplets before it becomes Invalid and disappears, you need to grab it. Just look for an Overflow and see if it's all Triplets.
Unfortunately, it seem those filled with Triplets are Arecibo Tasks....oh well.



Hi,

A noise bomb has so many signals that it makes a standard SW and a special SW hard to decide when to stop recording them.
The WU is processed from start to finish and the collecting part of the software has to have a limit what to store in CPU RAM.
One round is to check signals that have been time-delayed or -forwarded and divided to lengths of 8, 16, 32, 64, 128, 256, 512, 1024, 2048, 4096, ....
Each round could provide hundreds or thousands of "found" signals.
Imagine that to be done in parallel.
Quite hard to select the "correct ones" since everything is "Correct" but only 30 can be reported.

A noise bomb is a noisebomb.
8) Message boards : Number crunching : Arecibo Overflow Tasks Run With CUDA Special App Sometimes Filled With Triplets Only (Message 1953297)
Posted 20 days ago by Profile petri33
Post:
One from me, sorry no fanout, it's gone ..
http://setiathome.berkeley.edu/workunit.php?wuid=3116515390


Thanks Brent for trying!
9) Message boards : Number crunching : Arecibo Overflow Tasks Run With CUDA Special App Sometimes Filled With Triplets Only (Message 1953295)
Posted 20 days ago by Profile petri33
Post:
// WAS 11 in V0.97 before b2
#define MAX_TRIPLETS_ABOVE_THRESHOLD 13
that is in pulsefind cuda code.

Someone a long time before me has decided a hard limit. 11 or 13 or some value is the maximum of number of ...
If I make that number bigger it slows down triplet finding but reports more of them. If I make that smaller it misses triplets.

Easy to fix. Make it 11. Miss the rest of them.
10) Message boards : Number crunching : Arecibo Overflow Tasks Run With CUDA Special App Sometimes Filled With Triplets Only (Message 1953292)
Posted 20 days ago by Profile petri33
Post:
A noise bomb is a noisebomb.

It filled with results (probable signal candidates).

It is about of reporting order.
Which one of them to stop recording 30 of them and
where to start printing/writing them (max 30) to a file to submit to the servers.

Every one of found signals from a noise bomb could be claimed "valid"!

Ones, Tens, Hundreds, Thousands, Tens of thousands ...

So No help.

Petri.
11) Message boards : Number crunching : Panic Mode On (113) Server Problems? (Message 1953288)
Posted 20 days ago by Profile petri33
Post:
The return rate has been climbing steadily. It is now over 142k from ~90k before the BLC11 tapes.


A 50% increase is something a (relatively) new release of Linux CUDA app can do..
I doubt that there are so many of us using Linux CUDA V0.97b2 yet to affect return rate.
12) Message boards : Number crunching : User achievements thread......... (Message 1953286)
Posted 20 days ago by Profile petri33
Post:
+1

Congrats!
13) Message boards : Number crunching : Setting up Linux to crunch CUDA90 and above for Windows users (Message 1953226)
Posted 21 days ago by Profile petri33
Post:
"bad arg: -pfl
bad arg: 512"

The flag -pfl N is not supported any more.
Please do not try it with UPPERCASE. It is still there so that I can test it on offline benchmarks and bug hunting.

Same applies to -pfe.

Petri
14) Message boards : Number crunching : User achievements thread......... (Message 1953138)
Posted 21 days ago by Profile petri33
Post:
Total credit 	281,231,580
Average credit 	402,162.39  <-<<--<<<----<<<<-----<<<<<------ One machine only
==========================
15) Message boards : Number crunching : Setting up Linux to crunch CUDA90 and above for Windows users (Message 1953104)
Posted 21 days ago by Profile petri33
Post:
Gentleman.

Let us be quiet for a while. You two are on the front edge of a "flame war" if not in one. I want everyone to step away for a couple of days. Please.

And if possible, get the offending posts hidden and/or deleted on WOW's website.

Tbar is doing a great service for us getting these CUDA8.0 and then some setup for use.

Ian&Steve are asking questions.

And both of you are running into the single biggest problem with this type of communication. It loses all the context and body cues.

Please stop and give it a rest.

Tom (on my knees, even though they hurt)


Thanks Tom!

I was about to say something in a LOUD VOICE. Not in anger, not in haste. Just something to make "two kids on a block to stop". Period. "."
16) Message boards : Number crunching : Don't know where it should go? Stick it here! (Message 1952938)
Posted 22 days ago by Profile petri33
Post:
Thanks for the update Petri. Wasn't aware that there would be a CUDA10 release in conjunction with the Turing cards. I always read the compute tests in Anandtech's gpu reviews.


Happy to deliver some news.
Cuda 10 will be available when the cards are too.
I'm not going to get it any sooner even though I'd like to.

--
Petri
17) Message boards : Number crunching : Don't know where it should go? Stick it here! (Message 1952936)
Posted 22 days ago by Profile petri33
Post:
When a motherboard offers me an auxiliary PCIe slot power plug, I always use it and plug in a cable. Just to reduce the strain the slots have to endure with 4 X 75w = 300w going through them. The PCIe aux power connectors are usually either 4 pin Molex or 6 pin PCIe connectors and their location varies all over the place on the motherboards that feature them.


it's also pretty uncommon to see a board with those extra PCIe power connectors. i've only really seen it on some EVGA boards. or mining motherboards.

i know this practice is uncommon here, but one advantage to hooking GPU's remotely via a USB3.0 type riser, is that you have to power the card's PCIe slot externally, taking all the power strain OFF the motherboard and directly from your PSU. so you wont burn up the board


Asus WS-series of mobos have them too.
18) Message boards : Number crunching : User achievements thread......... (Message 1952934)
Posted 22 days ago by Profile petri33
Post:
My APR has dropped by 1000 but the credits per WU have come to the new high with blc-tasks. They pay now like 80 cr per WU compared to the old 50 / WU.
I do not know what happened.
Execution times have risen from 27-34 secons to 40+ seconds on my Volta but they pay well.
And my recent average credit seems to be rising. That may be due to version change -- less inconclusives.
19) Message boards : Number crunching : Setting up Linux to crunch CUDA90 and above for Windows users (Message 1952926)
Posted 22 days ago by Profile petri33
Post:
Well, Libcurl3 deals with networking. boinc will not function without it working correctly. I'm not surprised to see you are having other networking troubles as well. The Enable Networking in Recovery Mode not working should have been a clue. I would have reinstalled the System yesterday if it were me. The next time just install the WebKit and leave the Libcurl3 as it is, see if that works. If you ever do get it working, the CUDA 9.1 App will work with the 390 driver, that's partly why I chose to use it. Anyone running the Repository 390 driver can use the new CUDA App.

BTW, the Newer Systems use GDM as the display manager, that's why stop lightdm doesn't work on that system. I think the new cmd is something like sudo service GDM stop. That's partly why I changed to using the Recovery mode instead of the console, you don't have to stop anything in Recovery mode.


I use init 3.
20) Message boards : Number crunching : Don't know where it should go? Stick it here! (Message 1952924)
Posted 22 days ago by Profile petri33
Post:
BENCHMARKS 20x0!!!

The new 20x0(Ti) should be tested by the AnandTech-team or by the Tom's Hardware Page or the highly valued Swedish or Canadian site with the latest Linux CUDA version and compared against GTX like 980, 1080, 1080Ti and The TITAN V. And the results should be published.

I'll supply a version for the tests when the CUDA10 for the 20x0 is released and I get a Gpu(20x0Ti)card for development purposes for free or a loan. I'm not going to buy one in the future.

I bought one Volta (3k€) and I'm more than happy to buy some (n=3) used ones from the net. (reason: Perf/W)

--
Petri


Next 20


 
©2018 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.