Posts by HAL9000


log in
1) Message boards : Number crunching : Building a 32 thread xeon system doesn't need to cost a lot (Message 1784043)
Posted 4 hours ago by Profile HAL9000
Hal, thanks for the advice, I will look into both the adapters I'd need to pull this off, and those cards you linked to. So this pretty much limits me to a single powered GPU, plus another like the bottom end 950 or 750.

Maybe it's because it is is the Z600, which is the little brother to the Z800, and has available a 1100w PSU. I have never seen the inside of that one, but I'm pretty sure that they have more GPU power options available, otherwise, why bother, eh? ;-)

It might be worth checking to see if the Z600 & Z800 use the same PSU design. You might be able to swap in the 850w or the 1100w PSU from a Z800.

EDIT: Scratch that. I just has a closer look at the images for both.
2) Message boards : Number crunching : Building a 32 thread xeon system doesn't need to cost a lot (Message 1784026)
Posted 5 hours ago by Profile HAL9000
Checked out that list, wow, that one's going to be going bye-bye. So, next question, off the top of your heads, do you have an opinion of what the fastest 6 pin GPU would be? It can be a 600 or 700 series, as well as the 900, though I believe the connector on my GTX950 dual fan model is a single 8 pin, so that would take an adapter I mentioned earlier. I'd like to turn this one up a notch from the standard builds I have been doing with the 750/950s that I have, but I know that the single 6 pin is going to be my limitation. Thanks for your thoughts/opinions.

You may be limited by the PSU in that system. However doing a quick check looks like those workstations came with a 650w 80Plus Bronze PSU. So you might just be OK.
A PCIe x16 slot, by spec, should provide 75w. Note: Sometimes motherboards are not designed to actually accommodate this to save money.
A 6-pin PCIe connector is rated for 75w & an 8-pin PCIe connector is rated for 150w.

So with a single 6-pin connector that would limit you to a GPU under 150w. On paper a GTX 970 is 145w, but all of the 970's I've come across have either an 8-pin or 8-pin + 6-pin connector. Some GTX 960's do have a single 6-pin connector. So that kind of limits your options in the 900 series without using any adapters.

If you wanted to go with adapters a GTX 980, 165w, with 2 6-pin connectors might be a good option. Like the 04G-P4-2981-KR or 04G-P4-2983-KR.
3) Message boards : Number crunching : Building a 32 thread xeon system doesn't need to cost a lot (Message 1783981)
Posted 6 hours ago by Profile HAL9000

My experience is per machine, not per CPU, at least on my HP Z600, which has 2x Xeon E5620's. I get 100 ea. for two GPUs, and 100 for both CPUs. Unfortunate.

FWIW, HP Z600s have been going pretty cheap on eBay lately. Mine was ~$285 with the above CPUs, 12 gb ram, a 128gb SSD, a 1tb HDD and Win7x64 pro installed and licensed. Room for 2x x16 dual slot GPUs. Pretty good way to get into the multi-CPU game with a turn-key solution.

Well Jim, your experience with the Z600 piqued my interest, so I happened across one on fleabay a week or so ago for what seemed like a very reasonable price. It came with a single Intel Xeon E5649 2.53GHz SLBZ8 12MB 5.86 GT/s LGA1366 6-Core Processor, so I went out and found a 2nd one for about $50 and the standard, not high performance version of the HP OEM cooler for another 60 or so. I installed it the other day, and was looking at the wiring inside there, in terms of video card capabilities, and noticed that it only had a single 6 pin power connector for the video off of the PSU, though there is a 2nd power cord for SATA or IDE HD's that could be used with an adapter to I believe make a 2nd 6 pin, but not sure if it's configured correctly, as those usually use 2 different cables from the PSU, and this one would be using 2 connectors on one cable, since that's all there is in there.

I was a little suprised, as I'd have thought that being a workstation, it would be setup for installing a couple high horsepower GPU's into it. This one came with an Nvidia Quadro FX1800 in it, which has no power connectors, so would I be correct in presuming that it is a very low end card?

Other than putting in whatever the fastest GPU that has a single 6 pin connector along with a 2nd card that doesn't require any external power, what is a person to do for Moar Power? lol Or, is it just "it is what it is"? Thanks, this should be interesting.

*edit* Do Quadro's even work with BOINC & SETI?

List of Nvidia graphics processing units
4) Message boards : Number crunching : Can I further optimize multiple GPU calculations? (Message 1783506)
Posted 2 days ago by Profile HAL9000
Raistmer, guys, after a few years of having it sit on the shelf while waiting for the prices of the processors I wanted to become more 'reasonable', I finally got the $ together and bought the ones I had planned to use since the beginning. It is an EVGA SR-2, with a pair of X5690 CPU's, (currently with, for testing purposes) 3 GTX950's and a GTX750 with 48 gb of RAM running on a pair of Intel 480Gb SSD's in a RAID 1 config. The system is ID: 7990258 - CrunchMonster.

I spent much of yesterday building, loading, configuring and updating it, and finally early this morning turned on the spigot and let the WU's flow! Just before that I spent a bit of time reading the current posts on how to configure this to provide the most efficient crunching and utilize the hardware optimally, including Can I further optimize multiple GPU calculations? and my post Getting back into it, advice appreciated, as well as this one.

I have set it to <count>0.5, and set the <avg_ncpus>0.35</avg_ncpus> and <max_ncpus>0.35</max_ncpus>, and was reading about polling as well, though haven't implimented it yet, don't want to change too much at once, before consulting the Oracle (the brain trust here on the forum!).

As currently configured, it is running 24 CPU tasks, which I am sad to say seem to be averaging over 3 hours, which if I remember correctly is Much, Much longer than my CAD system I just built (ID: 7949285 Zoom-PC). Taking a quick look at that one, it appears that GPU tasks have a Run time (sec) averaging 750-1500 and the CPU time (sec) of 175-320 and the CPU tasks Run time is 5400-5700 and CPU time 5100-5500 (with a few in the 7500 range).

For the new system, (so far, as only 21 have validated as I write this) GPU tasks average Run time 1500-2500 & CPU time of 460-680, and CPU tasks Run time 11,000-12,500 CPU time of 9,700-10300! I know there is a few gen difference between the CPU's, and also between the 980's/970 in the former and the 950's/750 in the latter, but it seems that there might be something more at play here? Could someone with better insight than I take a look at them and see if there is anything that might stand out as a configuration Gotcha that I can adjust to help optimize this new one for a little better performance? The video cards are slightly O/C'd (around 10%), otherwise everything else is currently running stock.

Thanks again, guys!

Looking at your host 7990258. You may want to reduce the number of CPU tasks. When I was running a dual Xeon E5645 system I found 20 threads of MB ran much better than 24. When we had a lot of VLARs running 24 of them at once made the system less responsive. That should also help your GPUs. As their CPU time seems high to me.

Currently with my 750 ti I am only using processpriority = high in my mbcuda.cfg. I did try pfblockspersm = 8 pfperiodsperlaunch = 200, but found those values were slower than default in my configuration. I figured I might giving the -polling option a try next to see how it fairs.
5) Message boards : Number crunching : not able to add Seti@home to my BOINC (Message 1783494)
Posted 2 days ago by Profile HAL9000
That makes perfect sense - by default a firewall should prevent programs access the outside world.

And the default firewall setting it to notify users that a program needs accesses and add it if they wish.
6) Message boards : Number crunching : Task Preference Not Being Honored (Message 1783491)
Posted 2 days ago by Profile HAL9000
I have the exact same preferences and aside from some messages about no work when there are only vlar, no problems really to fill my cache. So it does work.

Perhaps the feeder was just empty or filled with VLARs for all of their previous requests.
7) Message boards : Number crunching : not able to add Seti@home to my BOINC (Message 1783291)
Posted 3 days ago by Profile HAL9000
2016-04-28 8:36:46 AM | | Fetching configuration file from http://setiathome.berkeley.edu/get_project_config.php
2016-04-28 8:36:58 AM | | Project communication failed: attempting access to reference site
2016-04-28 8:37:00 AM | | Internet access OK - project servers may be temporarily down.


I have been trying for weeks! any thoughts?

Can you open that page on the machine that is reporting the message?
http://setiathome.berkeley.edu/get_project_config.php

If not does http://setiathome.berkeley.edu/ open on that machine?
8) Message boards : Number crunching : Windows 10 - Yea or Nay? (Message 1783106)
Posted 3 days ago by Profile HAL9000
http://www.pcworld.com/article/3028541/privacy/tired-of-telemarketers-now-you-can-turn-the-tables-on-them-with-this-clever-bot.html

http://www.pcworld.com/article/206972/How_to_Sue_Telemarketers_and_Win.html

Every now and then I will get a robocall, but since I have my number registered with https://www.donotcall.gov/. I just report the number when I do get one.

You do know that there are some call centers that purchase the do not call list as a list of numbers to call. Mostly the ones that are offshore and thus don't have to obey the list.

Of course they do. It's they same thing teens did in the 90's with the apps that blocked the "naughty sites". Where they would d/l the software to get a list of sites to go to.
The legitimate call centers are normally not the problem anyway, but the last call I revived was from one. Last year there was some kind of vote going on regarding something with electric utility providers. However I didn't have a clue about the issue as my mobile number is from a state where I have not lived in over 10 years. Mobile lines, unlike land lines, are not permitted to have political robocalls.
When I still had a land line I would answer calls that had no Caller ID info like a business. "Thank you for calling XYZ corp. How may I direct your call?" The phone system had different voicemail boxes. So I had my sister record a greeting directing callers to "Press 1 for dept A, 2 for dept B" and so on. It resulted in some rather amusing messages, My favorite was someone calling about a job because their cousin worked there.
9) Message boards : Number crunching : Windows 10 - Yea or Nay? (Message 1783060)
Posted 3 days ago by Profile HAL9000
http://www.pcworld.com/article/3028541/privacy/tired-of-telemarketers-now-you-can-turn-the-tables-on-them-with-this-clever-bot.html

http://www.pcworld.com/article/206972/How_to_Sue_Telemarketers_and_Win.html

Every now and then I will get a robocall, but since I have my number registered with https://www.donotcall.gov/. I just report the number when I do get one.
10) Message boards : Number crunching : Bitcoin Utopia (Message 1782821)
Posted 4 days ago by Profile HAL9000
I think I mentioned it on the BCU fora after the first SETI@home campaign ended.
I thought it would be nice to have either a yearly or quarterly SETI@home campaign going.
I'm not sure why there hasn't been other one since the first. It could be lack of time for the admins to manage whatever is required for a BCU campaign, or it could be a political issue from the campus.
11) Message boards : Number crunching : Lunatics Help (Message 1782819)
Posted 4 days ago by Profile HAL9000
Does anyone know if there is a way to prevent Astropulse from running on the cpu and only run on the gpu?

Thanks,

Allen



. . I am only a rank beginner but if you run Lunatics and in the setup select Astropulse only for the GPU and not in the CPU section that should do it. Otherwise there is the app_config.xml or app_info.xml files. I have been warned not to fiddle with app_info.xml unless you are very sure of what you are doing, so app_config.xml might be your best bet.


. . HAL 9000 kindly provided this example :




Here are some examples of app_config.xml settings.
Running 1 task per GPU & reserving 1 CPU core per GPU task.
<app>
<name>setiathome_v8</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
For a system with 1 GPUs: 1 GPU tasks would run & 1 CPU cores would be reserved.
For a system with 2 GPUs: 2 GPU tasks would run & 2 CPU cores would be reserved.



. . These are for the GPU tasks but I think if you use the

<cpu_versions>

</cpu-version>

. . as brackets it will relate to what the CPU handles. But use the application name for the Astropulse not Seti@homeV8

. . Something like this ....

<app_version>
<app_name>astropulse_v7</app_name>
<cpu_versions>
<avg_ncpus>0.0</avg_ncpus>
</cpu_versions>
</app_version>

. . If I have that wrong hopefully somebody will correct me.

You might want to have a look at the app_config.xml documentation again. <cpu_versions> is not a valid option.

The easiest way to select which apps run on specific devices is to select them in the lunatics installer.


If Astropulse is unchecked while selecting CPU apps then an app_info.xml will be built without a CPU Astropulse app. The same goes for GPU apps. Select everything you want the device to run.
In the image I provided only AP was selected for the GPUs. If you wish to run MB apps on the GPU then you would need to select MB & AP applications for the GPU. As the installer is going to build a new app_info.xml instead of modifying your current one.
12) Message boards : Number crunching : Panic Mode On (102) Server Problems? (Message 1782678)
Posted 5 days ago by Profile HAL9000
I looked at your last AP and it shows the clock rate as 1019, which is about where it should be. I then went to the same time frame as that AP and found the cuda tasks around a "normal" clock rate. Looking ahead and backwards from that point it appears that when the machine is restarted it uses a Different clock rate. It will use that rate until it is again restarted. I looked at other 780s and noticed their rate was pretty consistent. So, the question is what is changing your clock rate after a reboot?

Looking as far back as possible it seems it was working fine with Version 8 CUDA;
Received 29 Feb 2016, 16:04:27 UTC, GPU current clockRate = 1123 MHz
1123/24 seems to be a consistent rate on a few machines.
This is where the trouble begins, Received 19 Apr 2016, 3:23:31 UTC, GPU current clockRate = 1215 MHz
That task began as 1123 and after a restart it was 1215.
It continued as 1215 until it was restarted here at 1097;
Received 21 Apr 2016, 19:35:13 UTC, GPU current clockRate = 1097 MHz
Then it worked fine until it was restarted here, https://setiathome.berkeley.edu/result.php?resultid=4879467266
Until the next restart it was bad news while clocked at 1201, https://setiathome.berkeley.edu/result.php?resultid=4879629574
Here it was restarted at 1136, https://setiathome.berkeley.edu/result.php?resultid=4883014586


now here is something i can go on ....
Indeed it is quiet weird why it would do that ...i wonder if precision is going heywire ... i checked everything kboost oberboost overvoltage etc everything is off and at default exept the fan it is running at 100% i'm gonna reboot it a couple times and see if that changes it and actually do a couple turn on then shutdown pull power from it and reinstall a clean driver

@Jason_gee
The early titan cards are based on the 780 chip so i bet if you get a good one you could ...gotta look at the asic

Based on the figures others have posted & what I am seeing with my 750 ti FTW running well over the spec'd clock rate. I would have to guess that EVGA has not implemented a clock cap for the boost. Likely only being throttled by thermal and voltage limits.
While my 750ti FTW is running at 1345MHz GPUz reports "GPU performance is being limited due to insufficient voltage". If you have a tool that allows you to adjust the voltage you could try lowering it until the clock is closer to the value in the specs.
13) Message boards : Number crunching : Lunatics Help (Message 1782669)
Posted 5 days ago by Profile HAL9000
I read the manual, they made a setup GUI since I installed it.

There are 2 important files:

remote_hosts.cfg .... put in the IP of the computer(s) you want to allow control

gui_rpc_auth.cofig .... put in a password

EDIT: and restart BOINC

For LAN connections I use the cc_config.xml option <allow_remote_gui_rpc> rather than using the remote_hosts.cfg.
14) Message boards : Number crunching : Average Credit Decreasing? (Message 1782427)
Posted 6 days ago by Profile HAL9000
Seems like over the last couple of days, my RAC has stabilized (on one machine) and even increased for two days (on the other), after falling for several days on both.

Is this just a fluke, or is the worst over?


I predict oscillation with work mix, around +/- 37% of some approximate midpoint. As oscillation instabilities work, some hosts will stay on the low side, and others stay on the high side of that, 5% in each end case never crossing the middle over extended periods, and the other 90% of us flopping around.

Prior to v8 & GBT work I a RAC oscillation of +/- 25-30% was not uncommon on some of my machines. The CPU only machines had a tendency to wiggle the least.
15) Message boards : Number crunching : Panic Mode On (102) Server Problems? (Message 1782409)
Posted 6 days ago by Profile HAL9000
It appears that the boys in the lab have done something to prevent downloads of (Nvidia) GPU WUs. If not all, then at least the GB ones and there are none available from Arecibo. Downloads have been working normal for me so I do not think I have done anything to prevent them from downloading. I have currently 100 CPU WUs in progress but only 80 GPU WUs despite asking for them.


Edit: Guess I called a "no hitter". Just received a couple of guppi GPU WUs. Not sure why they do not seem to be available in quantity though.

As we know VLAR tasks are not sent to GPUs. GBT data is expected to be mostly VLARs. If we are on a data sets from Arecibo that generate mostly VLARs as well. Then there will be few tasks for GPUs.
16) Message boards : Number crunching : Something's happenin' here, what it is ain't exactly clear.. (Message 1782250)
Posted 6 days ago by Profile HAL9000
Gotcha, crunching away! Thanks for the quick reply.

We don't know what tasks are overflows until we try to crunch them. At least they go quickly.
17) Message boards : Number crunching : Panic Mode On (102) Server Problems? (Message 1782245)
Posted 6 days ago by Profile HAL9000
All i can tell you is that none of the cards are overclocked they all on default


Something's odd then.

The defaults for the reference GTX780 are
Base clock: 863MHz
Boost clock: 900 MHz

The EVGA 780 SC being a factory overclock are
Base Clock: 967 MHZ
Boost Clock: 1020 MHz

So something has boosted your clock speed if it is running at 1201 MHz
PrecisionX settings? PCIe bus overclock?

EDIT-
Beaten by TBar.
What does GPUz have to say?

The GTX 750 Ti FTW I have does some kind of self overclocking. It is meant to run at 1189MHz and have a boost of 1268MHz. However once the SETI@home app starts it clocks up from 300MHz to 1189MHz. Then a few seconds later goes to 1345MHz and stays there. Since it hasn't spit out any errors yet I haven't worried about it.

The host is a new install of Win7 & only has BOINC, the Nvidia driver, & monitoring apps, like CPUz/GPUz, on it
18) Message boards : Number crunching : Panic Mode On (102) Server Problems? (Message 1782105)
Posted 7 days ago by Profile HAL9000
Yea my overflows have definitely increased significantly :(

I would rather be re-crunching the 2010 MB data than proving that the GBT data process needs work.

I would say to limit 1 GBT file at a time until they figure the problems out, because it really looks like we will have to re-crunch them again.

Well, either the *MESSIER031* tasks are broken, or they have found a radio transmission from a very old rock consert, from some planet in the Andromeda galaxy :-)

These *MESSIER031* tasks are truly noisy as a rock concert.

The nature of the data we see from GBT could be much different from that we have typically seen from Arecibo.

It could be that something is causing the applications to trip up & consider the data to be noisy.
Or we might be looking at data from pulsar observations.

Out of the tasks I am seeing on my 750ti about 2/3 result in an overflow. Is that a lot? I don't know. ¯\_(º¿º)_/¯ Many of them are coming up valid against stock CPU apps & not just other GPU apps. So that gives me a feeling that they are legitimate overflows. Unless as I said there is an error in the apps, but I feel something that major would likely have been found at Beta.
19) Message boards : Number crunching : Average Credit Decreasing? (Message 1781950)
Posted 7 days ago by Profile HAL9000
Dr. Anderson USUSALLY only responds when you can prove that one plus one does NOT any longer equal two plus or minus ten digits either way.
The outcome creditwise usually does not concern him much because credits are not a scientific matter, and are naturally not part of his response curve.
So somebody complaining that credits....HIS credit system....are broken naturally just slide off of his back into nowhere land.

It has to take a LANDSLIDE of massively wrong (or correct, in this case) information to sway the man one way or another.
We all have to continue to speak up for justice in this matter until the ship is righted.

Meow.

One issue that software faces is how people perceive how it should work. Even if CreditNew is working 100% as designed it is perceived as broken.
As credit is a means to determine performance, or contribution, to a project it seems counter intuitive to me for that to be highly variable. If I configure a machine to run a project a fixed amount of time a day. It would be 1 hour a day or 24/7. It makes logical sense that it should be contributing a fairly constant amount of processing power and by extension generate a fairly constant credit value.
It make it hard to figure out what is going on when the reference isn't a consistent value.
20) Message boards : Number crunching : GPU WU crunching (Message 1781581)
Posted 8 days ago by Profile HAL9000
Forget about Core(TM)2 procs...they don't have intelHD!

write which intelHD xxxx you have in i3 or i5 proc?!

The i3-2350M has Intel HD Graphics 3000. As I mentioned earlier in this thread it is to old to support OpenCL.


Next 20

Copyright © 2016 University of California