Well I done it. I now have a 24 "core" Boinc server online.

And due to help of the other Boinc/Seti Volunteers I have been able to constraint World Community Grid to "only" 4 cores. Rosetta to "only" 2 cores, Citizen Science Grid to "only" 2 cores which allows the Scheduler full run of the rest of the cores for SetiBeta and Seti (of course).

Now I know how an owner of a 32 core machine "feels". Just looking at those cores displayed in Task Manager :D

Tom

ps, Yes, keeping it cool enough has raised its head again :(

Congratulations!

---

To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones

Congratz! Fun looking at task manager, as well as SETI manager, seeing all those tasks running, huh?

---

Congratz! Fun looking at task manager, as well as SETI manager, seeing all those tasks running, huh?

<G I A N T G R I N>


Tom

---

A proud member of the OFA (Old Farts Association).

Early this morning, I figured what the heck, and ordered a few of them from Newegg, they were cheap enough, the suckage is that they are coming from CA, and using Super Eggsaver (4-7 bus. days) shipping, it means that I am actually probably not going to see them till the following Monday..

Al,
Its Thursday evening. So you should have gotten the power supply adaptors to put in two Gtx 1060's in your Z600 by now.

How is it going? Are things "smoking along" (as it it went up in a puff of) or is everything "cool"? :)

So far, I'm "cool". Or at least my Z600 is.

I was reminded that things don't always fail "fast". I had an after market PS (500 watts)[standard PS is 235 watts) and a Gtx 1060Ti (compact) in a Dell OptiPlex 7010 Microtower. It came up and ran "fine" for a month. Then when I took the system down to swap the 1060 into the Z600, the MB of the Dell 7010 reported "I fried" (2 blinks, pause, 1 blink, amber power up button translates into "replace system board"). I have replaced the MB and reverted to the original PS and a Gtx 750Ti and it is humming along (a little slower though :)

Tom

---

A proud member of the OFA (Old Farts Association).

Hey Tom, they arrived yesterday, and I am in the process of firing it up and doing necessary driver updates right now. I just got done uninstalling the old video drivers I had installed a year ago for I believe the 700 series cards that were in it, and am putting on the latest 384.76 drivers on it using clean install. Still trying to understand why it feels it needs to install MS Visual C++ Redistrib packages (both x86 and x 64 versions to boot)during a video driver install, but I removed them once the driver was installed properly. Just more crap to gum up the works on a dedicated cruncher, and more virus vectors to exploit too methinks. It isn't crunching yet, but other than it beeping 4 times (continuously, with the power button red colored) the first time I powered it up, which went away when I hit the power button again, it seems to be working ok so far, at least during the software uninstalls and installs. Any thoughts on what those signified would be appreciated, as I haven't fully powered it off yet, just soft reboots between driver installs.

For a reference, I am running 2 E5649 procs in it, not sure why I went with those, usually on these older machines I like to stretch the envelope to the fastest procs available in the series, but it might still have been about the time when Mac people were bidding up the top end versions of these for their Mac Pro's or something. Oh well, we'll see how it does at a reported 2.53GHz. Just installed Precision X OC, and the damn thing is trying to install .net, even though I did the trick I learned from a tech at EVGA and nuked any server named files in windows explorer before saying OK to it completing, which supposedly eliminates any need for anything .net. After a reboot while it was trying to install them, it loaded up and seems to be running just fine.

The latest Precision has some interesting sensor monitoring abilities with the iCX card, more than I remember seeing with whatever version I have installed on the system with the 980Ti's and the 1080, not sure if it is iCX capable card related, or if it is just that it's the latest version of it has more monitoring built in. Oh, and here is the event log of this machine, I'm pretty sure that it doesn't have the latest of anything on it, as I just loaded it up for testing in mid May last year, then powered it down till today:

7/12/2017 8:21:36 PM | | Starting BOINC client version 7.6.22 for windows_x86_64
7/12/2017 8:21:36 PM | | log flags: file_xfer, sched_ops, task
7/12/2017 8:21:36 PM | | Libraries: libcurl/7.45.0 OpenSSL/1.0.2d zlib/1.2.8
7/12/2017 8:21:36 PM | | Data directory: C:\ProgramData\BOINC
7/12/2017 8:21:36 PM | | Running under account user
7/12/2017 8:21:38 PM | | CUDA: NVIDIA GPU 0: GeForce GTX 1060 6GB (driver version 384.76, CUDA version 9.0, compute capability 6.1, 4096MB, 3044MB available, 4762 GFLOPS peak)
7/12/2017 8:21:38 PM | | CUDA: NVIDIA GPU 1: GeForce GTX 1060 6GB (driver version 384.76, CUDA version 9.0, compute capability 6.1, 4096MB, 3044MB available, 4762 GFLOPS peak)
7/12/2017 8:21:38 PM | | OpenCL: NVIDIA GPU 0: GeForce GTX 1060 6GB (driver version 384.76, device version OpenCL 1.2 CUDA, 6144MB, 3044MB available, 4762 GFLOPS peak)
7/12/2017 8:21:38 PM | | OpenCL: NVIDIA GPU 1: GeForce GTX 1060 6GB (driver version 384.76, device version OpenCL 1.2 CUDA, 6144MB, 3044MB available, 4762 GFLOPS peak)
7/12/2017 8:21:38 PM | SETI@home | Found app_info.xml; using anonymous platform
7/12/2017 8:21:38 PM | | Host name: ZeeCruncher
7/12/2017 8:21:38 PM | | Processor: 24 GenuineIntel Intel(R) Xeon(R) CPU E5649 @ 2.53GHz [Family 6 Model 44 Stepping 2]
7/12/2017 8:21:38 PM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 sse4_1 sse4_2 popcnt aes syscall nx lm vmx smx tm2 dca pbe
7/12/2017 8:21:38 PM | | OS: Microsoft Windows 7: Professional x64 Edition, Service Pack 1, (06.01.7601.00)
7/12/2017 8:21:38 PM | | Memory: 47.98 GB physical, 95.97 GB virtual
7/12/2017 8:21:38 PM | | Disk: 1.82 TB total, 1.71 TB free
7/12/2017 8:21:38 PM | | Local time is UTC -5 hours
7/12/2017 8:21:38 PM | | Config: event log limit disabled
7/12/2017 8:21:38 PM | | Config: use all coprocessors
7/12/2017 8:21:38 PM | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 7995769; resource share 100
7/12/2017 8:21:43 PM | SETI@home | General prefs: from SETI@home (last modified 03-Apr-2013 23:59:56)
7/12/2017 8:21:43 PM | SETI@home | Computer location: home
7/12/2017 8:21:43 PM | SETI@home | General prefs: no separate prefs for home; using your defaults
7/12/2017 8:21:43 PM | | Preferences:
7/12/2017 8:21:43 PM | | max memory usage when active: 24567.61MB
7/12/2017 8:21:43 PM | | max memory usage when idle: 46678.46MB
7/12/2017 8:21:43 PM | | max disk usage: 100.00GB
7/12/2017 8:21:43 PM | | (to change preferences, visit a project web site or select Preferences in the Manager)
7/12/2017 8:27:14 PM | | Suspending computation - user request


Any suggestions for updates to it that don't get _too_ in depth? I NNT'ed it before I shut it down last year, but I am thinking of just allowing tasks and seeing how it goes. Thoughts on that one with the info provided above, good idea or wait till things are updated? Oh, and not trying to take over your thread, Tom, but you asked... ;-p lol

*edit* Meh, the daughter is bored being here at the shop, so sadly I will have to head out in a couple, I don't feel comfy leaving this thing running without me being here to supervise it, esp with the power 'modifications' I've done to power those cards, so I'm going to shut it down for now, but if you have thoughts on updates I might want to do before I let it start crunching again, let me know. Or should it be good enough to go as is? At least it seems to be working, and with the front cover off, and nothing crunching, the CPU temps are between 36-40*C, and the GPUs are running unloaded at 32. One thing that will be interesting is how they react to being loaded, and due to the MB design, being they are so close together, as well as not being blower cards, so they toss the heat back into the case. Will be an interesting test to see how they react to heavy loads. Well, signing off for now, will try to resume this adventure tomorrow sometime.

---

Hey Tom, they arrived yesterday....For a reference, I am running 2 E5649 procs in it, not sure why I went with those, .....Any suggestions for updates to it that don't get _too_ in depth? I NNT'ed it before I shut it down last year, but I am thinking of just allowing tasks and seeing how it goes. Thoughts on that one with the info provided above, good idea or wait till things are updated? Oh, and not trying to take over your thread, Tom, but you asked... ;-p lol

Yes, I asked.

I started out with an single E-5620 (2.4Ghz). I have the c2 revision of the motherboard (some 2010 date in the boot block). I tried every proposal for getting an X5680 installed. None of the ideas worked. The X5675 (6c/12ht) 3.06Ghz does work but I paid $100 for the pair. If you are mainly going to drive Seti on the gpus and run Seti and "whatever" on the cpu's, it may not be worth the money to upgrade past the E5649's. Those are working. The X56xx's will only work in the C2 version of the motherboard. Since the cpus you have are 6c/12t and MIGHT (80 watt vs. 95 for the X56xx) run cooler I would stand pat for now.

1) During a HD failure I attempted to install Windows 7 w/o SP1 and it said "I can't do that".
2) I was able to "migrate" my working Windows 7 SP1 to it.
3) I expect yours to get very "hot" under load, which is why I have yet another set of cheap(er) cpu heatsinks on order. If these work, I will post the specs.
4) How hot? 75C with bio fan setting at about 4 (just below the whine).
5) Temperature as reported by TThortle.

I am running it that hot inspite of the Intel spec that say it should not run above 71C because I have an i7 from Dell where Dell support said 75C was a "normal" operating temperature for a heavily loaded 7010. Since the cpu shutdown is north of 90C I am only worried, not shutdown (yet).

If I am understanding you correctly you were asking if there is anything I have done to alter all the Windows 7 updates that come down after you install Win7 from scratch. Not really. I do have a copy of SP1 since my best starting Win7 image doesn't have it. I also have the two standalone patches that fix the win7 update delay issue if/when I run into it. I haven't made any attempts to get rid of anything that Nvidia, Xoc or TThortle haven't wanted to install although I did do a custom install so the GeForce "experience", and the 3D stuff were not installed from Nvidia.

If that is not what you were asking about I will admit to running Lunatics beta6 with two modifications. (a later gpu app from Mikes World and hand installed SetiGraphics back in[Thank you Jord/Ageless]).

Since in Win7 at least it considers video drivers "optional" I have not had to fight with MS about keeping a working version of the video drivers.

I suppose we could spin off a Z600 thread but the moderators once noted that I have "spawned" an awful lot of different threads so maybe not.... :)

I hope your 1060's include one full length and one compact since that will allow better fan clearance. Otherwise just turn up the gpu fans higher. ;)

Other suggestions for keeping it "cool" are sit it out under a shady tree in the Minnesota cool up there.... (nah, wait till fall).

Hope the daughter is happier after her Daddy maintenance stop.

Tom

---

A proud member of the OFA (Old Farts Association).

Thanks for the heads up on your setup. My thoughts on updates were regarding the version of BOINC, and the additional apps related to it. On a related side note, I had taken a look at the updates in Windows that it had wanted to do, there were I think 21 of them, plus a number more recommended ones, but didn't think I had scheduled any to actually install, I was just taking a look see. Sadly, when I went to shut it down, it started installing the 21 and there was nothing I could do to stop them at that point. Grrr. I am going to see what all happened when I boot it up today, hopefully everything will be the same after all that stuff is installed and I don't get any nag notifications from Windows.

Does what I posted in my startup log indicate anything I should do, update any apps, or doesn't it show enough info to make a recommendation? If not, let me know what info I should gather once I get back to the shop and get it booted up. I'd be interested in the mods you made to yours, and which mods you did. As to the card length, they are both full length cards, and it just _barely_ clears the 5 1/4" CD drive bay, it is pretty tight in there. I can only imagine that under full load, it is going to get pretty toasty in there with the cover on. Might be a candidate for side cover off/box fan cooling, as I remember last time I had it crunching when I just built it, with the 700 series cards, it was some Hot air coming out the back of it...

You could probably migrate this into a Z600 or even just a Z thread, as I've seen a few 800s that are coming down in price, as well as the newer Z series (20/40) as they come more reasonable in the price in the future. Could be a good reference for those interested in these machines, and might keep the mods happier with you. ;-) Maybe they could rename the tread to be more appropriate to that end if you do decide to go in that direction.

---

You could probably migrate this into a Z600 or even just a Z thread, as I've seen a few 800s that are coming down in price, as well as the newer Z series (20/40) as they come more reasonable in the price in the future. Could be a good reference for those interested in these machines, and might keep the mods happier with you. ;-) Maybe they could rename the tread to be more appropriate to that end if you do decide to go in that direction.

I like the idea of migrating these a "Z-X00" aka: Z400/Z600/Z800 thread. But I am ignorant on how to do that. I assume I need to get in touch with a moderator?

I know I could start such a thread and then copy/paste all the relevant messages but it might make "them" unhappy? I'm going to try for an email to see what they have to say.

---

A proud member of the OFA (Old Farts Association).

Thanks for the heads up on your setup. My thoughts on updates were regarding the version of BOINC, and the additional apps related to it....

Does what I posted in my startup log indicate anything I should do, update any apps, or doesn't it show enough info to make a recommendation?....t is pretty tight in there. I can only imagine that under full load, it is going to get pretty toasty in there with the cover on. Might be a candidate for side cover off/box fan cooling, as I remember last time I had it crunching when I just built it, with the 700 series cards, it was some Hot air coming out the back of it...

1) The next time you boot Win7 it will "finish" installing those 20 odd updates before it will let you logon. I would be very hesitant to bother the updating that would currently be in progress. I have had luck picking/choosing about what I wanted but you need to clear all the check boxes before you say "ok" or it will take that as permission to install those updates.

2) I have re-installed newer version of the Boinc client download over older versions while Boinc was not running and never lost any WU's. I don't think I lost any app_config.xml files or the mb*sog.txt files either. Not sure about the app_info.xml files.

3) Unless you are willing to tinker you will probably lose all the WU's you have on your HD if/when you upgrade to Lunatics Beta6. So either be willing to run the tasks down to zero. Or be willing to modify the "app_info.xml" file that is generated by the Beta6 installer to include the current cpu and gpu version numbers before you restart Boinc. (see "wu missing" in the messages/number crunching message area) or I can post my copy of the "how to that I got".

4) The "newer" gpu app can be downloaded from "Mikes World" and then use notepad to replace all the instances of the old app with the new app name. It is not supposed to be any faster but it produces fewer computation errors. See discussion in Crunching message area on "motivating yourself to move from rxxxxx to rxxxx".

5) My own limited experiment with leaving the cover off my Z600 didn't show any difference in how "hot" it was running. I think my latest non-$200 cpu heatsinks will show up tomorrow. (They are both LGA 1366 and "short" enough they should allow the cover to be replaced). The question is how much butchering will I have to do to get the power plugs to fit? At $44 for two and come highly recommended in the newegg product review I am hoping to lose 10C by using them. That would take me from about 75C down to 65C which is below the standard 71C Xeon limit.

5) To add SetiGraphics back into the Lunatics Beta6 you can follow the directions listed in the top of the "wishlist" message area. Or I can re-post them later.

6) To control the number of WU's that will run on a project by project basis you can use the following in individual app_config.xml files in each product.

An example for how to constraint "World Community Grid" follows:

<app_config>
<app>
<name>zika</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>mcm1</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>oet1</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>fahv</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>scc1</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>fahb</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>hst1</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<project_max_concurrent>4</project_max_concurrent>
</app_config>
End of code.  Everything below this is considered to be comments by Boinc.

Example for limiting for a single project/task within the app_config.xml
<name>zika</name>
<max_concurrent>2</max_concurrent>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>

Place this after the last </app> to limit the total instances running under that app_config.xml
</app>
<project_max_concurrent>4</project_max_concurrent>
</app_config>

One way to "find" the current project name is to put an incorrect project name in an app_config.xml file in a project directory. When you re-read the config files BOINC will tell you the "real names" of apps are currently in that subdirectory. You can also find the complete listing of project program names in the Boinc directory "client status...." files.

Boinc will complain about any of the currently missing WCG projects/programs. And does so routinely. I have used this on every project except Seti and they all read/respond to the task limitation parameters.

Please note, I have found that controlling the number of specific running tasks for each project will not necessarily cause the scheduler to download more seti tasks, so it is possible to be running a small number of cores using the above parameters, until the Boinc "cache" gets small enough that the local scheduler is now convinced it wants to download more Seti cpu tasks (saw this behavior by watching the event log when I tried to get it to download more seti tasks).

HTH,
Tom

---

A proud member of the OFA (Old Farts Association).

In this thread https://setiathome.berkeley.edu/forum_thread.php?id=81586 the discussion has evolved from the topic of "Getting the most production for the least electricity" to getting high core count Z series machines setup for Boinc and kept cool.

Al proposed I/We start a thread devoted just to that. Now according to the "lists" of cpu's there are probably more than 500 Boinc users running HP Z400's so this is YOUR thread.

HTH,
Tom

---

A proud member of the OFA (Old Farts Association).

What started me down the path of Z machines and/or upgrades is the fact that I have a Z-400 with 6 memory slots. This means the machine will also run the X56xx series of cpu's as well as the W35xx series.

So I located both an W3580 (2nd fastest available in that series) for $45 from China and a X5680 (2nd fastest available in that series) 3.33Ghz with 6 cores w/HT which gave me effectively 12 cpus.

The X5690 cost twice as much. While the price has dropped significantly on both on eBay, the spread has not. So I now have a Z-400 running 12 cores of Boinc/Seti tasks.

The HP support forums offered speculation that it "should" be possible to run 2 X5680's or 90's on a Z-600. You need a 2010 Boot block date, the latest bio's update and a "high performance" cpu heat sink before it would work.

A high performance heat sink has a jumper from pin 1 to pin 5 and the Z600 one's are VERY expensive. However I did have a Z400 high performance cpu heat sink and a 2nd X5680 cpu so I set out to try this out on a Z600.

The long and short of it is, inspite of the Z600 allowing for up to a 130 watt cpu in its specifications it will NOT :( boot with a Z5680 cpu. The bios displays a "cpu draws to much power" message. Rats.

So I am now running my Z600 with the X5675 (3.06Ghz, 6 c/12t) and trying to find a way to get it to run cooler than 75C without spending $300 on "high performance" cpu heatsink/coolers. I have a new set of lower cost possibles on the way. Will post results here.

Tom

---

A proud member of the OFA (Old Farts Association).

Oh, yes. The upgrade path for the Z400 included getting a LCS (Liquid Cooling Solution) installed so I could run the cpu flat out while maintaining the cpu temperature at/below 71C.

---

A proud member of the OFA (Old Farts Association).

Thanks for this thread. I'll sure be watching and will contribute as I can.
The Z400 and even more the Z600 can be bargains to get, but challenging to deal with.

---

...The Z400 and even more the Z600 can be bargains to get, but challenging to deal with.

Here here! They are a somewhat closed/proprietary system, with little methods it seems of bending them to your will, if you have the correct blend of hardware and firmware. This will be an interesting thread as it goes along, esp as the prices of the older models continue to drop.

---

From the other thread:

Thanks for the heads up on your setup. My thoughts on updates were regarding the version of BOINC, and the additional apps related to it....

Does what I posted in my startup log indicate anything I should do, update any apps, or doesn't it show enough info to make a recommendation?....t is pretty tight in there. I can only imagine that under full load, it is going to get pretty toasty in there with the cover on. Might be a candidate for side cover off/box fan cooling, as I remember last time I had it crunching when I just built it, with the 700 series cards, it was some Hot air coming out the back of it...

1) The next time you boot Win7 it will "finish" installing those 20 odd updates before it will let you logon. I would be very hesitant to bother the updating that would currently be in progress. I have had luck picking/choosing about what I wanted but you need to clear all the check boxes before you say "ok" or it will take that as permission to install those updates.

2) I have re-installed newer version of the Boinc client download over older versions while Boinc was not running and never lost any WU's. I don't think I lost any app_config.xml files or the mb*sog.txt files either. Not sure about the app_info.xml files.

3) Unless you are willing to tinker you will probably lose all the WU's you have on your HD if/when you upgrade to Lunatics Beta6. So either be willing to run the tasks down to zero. Or be willing to modify the "app_info.xml" file that is generated by the Beta6 installer to include the current cpu and gpu version numbers before you restart Boinc. (see "wu missing" in the messages/number crunching message area) or I can post my copy of the "how to that I got".

4) The "newer" gpu app can be downloaded from "Mikes World" and then use notepad to replace all the instances of the old app with the new app name. It is not supposed to be any faster but it produces fewer computation errors. See discussion in Crunching message area on "motivating yourself to move from rxxxxx to rxxxx".

5) My own limited experiment with leaving the cover off my Z600 didn't show any difference in how "hot" it was running. I think my latest non-$200 cpu heatsinks will show up tomorrow. (They are both LGA 1366 and "short" enough they should allow the cover to be replaced). The question is how much butchering will I have to do to get the power plugs to fit? At $44 for two and come highly recommended in the newegg product review I am hoping to lose 10C by using them. That would take me from about 75C down to 65C which is below the standard 71C Xeon limit.

5) To add SetiGraphics back into the Lunatics Beta6 you can follow the directions listed in the top of the "wishlist" message area. Or I can re-post them later.

6) To control the number of WU's that will run on a project by project basis you can use the following in individual app_config.xml files in each product.

An example for how to constraint "World Community Grid" follows:

<app_config>
<app>
<name>zika</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>mcm1</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>oet1</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>fahv</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>scc1</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>fahb</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<app>
<name>hst1</name>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
<project_max_concurrent>4</project_max_concurrent>
</app_config>
End of code.  Everything below this is considered to be comments by Boinc.

Example for limiting for a single project/task within the app_config.xml
<name>zika</name>
<max_concurrent>2</max_concurrent>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>

Place this after the last </app> to limit the total instances running under that app_config.xml
</app>
<project_max_concurrent>4</project_max_concurrent>
</app_config>

One way to "find" the current project name is to put an incorrect project name in an app_config.xml file in a project directory. When you re-read the config files BOINC will tell you the "real names" of apps are currently in that subdirectory. You can also find the complete listing of project program names in the Boinc directory "client status...." files.

Boinc will complain about any of the currently missing WCG projects/programs. And does so routinely. I have used this on every project except Seti and they all read/respond to the task limitation parameters.

Please note, I have found that controlling the number of specific running tasks for each project will not necessarily cause the scheduler to download more seti tasks, so it is possible to be running a small number of cores using the above parameters, until the Boinc "cache" gets small enough that the local scheduler is now convinced it wants to download more Seti cpu tasks (saw this behavior by watching the event log when I tried to get it to download more seti tasks).

HTH,
Tom

Well, good news is that I have 0 tasks in it right now, so I can start dissecting it at any time. I'll take a good read thru this info once I get to the shop and can be sitting right in front of it, instead of trying to go my memory. That wouldn't be good... ;-)

---

I installed the latest version of BOINC, and then went to Mikes Worlds and got the beta 6 SOG and installed it too. Restarted BOINC and here is my event log:

7/14/2017 4:42:38 PM | | Starting BOINC client version 7.6.33 for windows_x86_64
7/14/2017 4:42:38 PM | | log flags: file_xfer, sched_ops, task
7/14/2017 4:42:38 PM | | Libraries: libcurl/7.47.1 OpenSSL/1.0.2g zlib/1.2.8
7/14/2017 4:42:38 PM | | Data directory: C:\ProgramData\BOINC
7/14/2017 4:42:38 PM | | Running under account user
7/14/2017 4:42:39 PM | | CUDA: NVIDIA GPU 0: GeForce GTX 1060 6GB (driver version 384.76, CUDA version 9.0, compute capability 6.1, 4096MB, 3044MB available, 4762 GFLOPS peak)
7/14/2017 4:42:39 PM | | CUDA: NVIDIA GPU 1: GeForce GTX 1060 6GB (driver version 384.76, CUDA version 9.0, compute capability 6.1, 4096MB, 3044MB available, 4762 GFLOPS peak)
7/14/2017 4:42:39 PM | | OpenCL: NVIDIA GPU 0: GeForce GTX 1060 6GB (driver version 384.76, device version OpenCL 1.2 CUDA, 6144MB, 3044MB available, 4762 GFLOPS peak)
7/14/2017 4:42:39 PM | | OpenCL: NVIDIA GPU 1: GeForce GTX 1060 6GB (driver version 384.76, device version OpenCL 1.2 CUDA, 6144MB, 3044MB available, 4762 GFLOPS peak)
7/14/2017 4:42:39 PM | SETI@home | Found app_info.xml; using anonymous platform
7/14/2017 4:42:39 PM | | Host name: ZeeCruncher
7/14/2017 4:42:39 PM | | Processor: 24 GenuineIntel Intel(R) Xeon(R) CPU E5649 @ 2.53GHz [Family 6 Model 44 Stepping 2]
7/14/2017 4:42:39 PM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 sse4_1 sse4_2 popcnt aes syscall nx lm vmx smx tm2 dca pbe
7/14/2017 4:42:39 PM | | OS: Microsoft Windows 7: Professional x64 Edition, Service Pack 1, (06.01.7601.00)
7/14/2017 4:42:39 PM | | Memory: 47.98 GB physical, 95.97 GB virtual
7/14/2017 4:42:39 PM | | Disk: 1.82 TB total, 1.71 TB free
7/14/2017 4:42:39 PM | | Local time is UTC -5 hours
7/14/2017 4:42:39 PM | | Config: event log limit disabled
7/14/2017 4:42:39 PM | | Config: use all coprocessors
7/14/2017 4:42:39 PM | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 7995769; resource share 100
7/14/2017 4:42:44 PM | SETI@home | General prefs: from SETI@home (last modified 03-Apr-2013 23:59:56)
7/14/2017 4:42:44 PM | SETI@home | Computer location: home
7/14/2017 4:42:44 PM | SETI@home | General prefs: no separate prefs for home; using your defaults
7/14/2017 4:42:44 PM | | Preferences:
7/14/2017 4:42:44 PM | | max memory usage when active: 24567.61MB
7/14/2017 4:42:44 PM | | max memory usage when idle: 46678.46MB
7/14/2017 4:42:44 PM | | max disk usage: 100.00GB
7/14/2017 4:42:44 PM | | (to change preferences, visit a project web site or select Preferences in the Manager)

Anything look unhappy, and what was it you mentioned about using notepad and find/replacing things? It's been a while since I have had to install this. Thanks!

Oh, and in case you hadn't noticed my sig, I'm a one project man. If SETI isn't feeding my rigs, they're just sitting around idling till they do. Been that way for over 18 years, I figure why change now? ;-)

---

I installed the latest version of BOINC, and then went to Mikes Worlds and got the beta 6 SOG and installed it too. Restarted BOINC and here is my event log:
------
Anything look unhappy, and what was it you mentioned about using notepad and find/replacing things? It's been a while since I have had to install this. Thanks!

Al,
There is a later version of the gpu SOG task available from Mike's World. r3557 is the current one. A slightly better (results) app is r3584. Download and unpack it into a staging location. Then copy everything but the mb*sog.txt file into the main Seti directory. You will want to keep your current mb*sog.txt so this won't delete it.

Probably would be safest to stop Boinc/Seti. Then use notepad to search and replace all r3557's with r3584 in "app_info.xml". Save. Restart Boinc/Seti. The near term result should be invisible. But here is an thread with an explanation

https://setiathome.berkeley.edu/forum_thread.php?id=81596 About what the r3584 does better.

There is a discussion on getting the most out of your Gtx 1060 at https://setiathome.berkeley.edu/forum_thread.php?id=81516

HTH,
Tom

---

A proud member of the OFA (Old Farts Association).

The long and short of it is, inspite of the Z600 allowing for up to a 130 watt cpu in its specifications it will NOT :( boot with a Z5680 cpu. The bios displays a "cpu draws to much power" message. Rats.

Tom

Edit: typo. That Z5680 should be X5680.

Tom

---

A proud member of the OFA (Old Farts Association).

,,, what was it you mentioned about using notepad and find/replacing things? It's been a while since I have had to install this. Thanks!

Al,
The other thing you can do with notepad is run the "SetiGraphics" app within the Lunatics distro just like stock Seti so you can see all the pretty graphs. :)

The trick is copying in the .exe file and then adding two entries into the app_info.xml file.

The how to is here: https://setiathome.berkeley.edu/forum_thread.php?id=81682

HTH,
Tom

---

A proud member of the OFA (Old Farts Association).

Now that is a little depressing. I was looking through the Top Hosts list when I ran into a W3690 which was reporting 12 processors. Was supporting 2 Gtx 1060's and had a RAC of 64,147.81 https://setiathome.berkeley.edu/show_host_detail.php?hostid=8201667

The W3690 cpu is one of the two fastest cpu's that will fit in a Z400 (the other is the X5690). And I have 2 Gtx 1060's in one Z600 box with twice as many cores (but mine run slower).

What's he got that I don't? Apparently its Linux and he might be running that "secret sauce" Gpu app :( The one that can't be ported directly to windows because of major differences in the OS video drivers.

As has been pointed out, there are always someone further out on the "outlier" than thou.

Tom

---

A proud member of the OFA (Old Farts Association).