Come on admit it, you have an E5-2670 crunching on Boinc and/or Seti@Home. How is it doing?

Message boards : Number crunching : Come on admit it, you have an E5-2670 crunching on Boinc and/or Seti@Home. How is it doing?
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 5 · 6 · 7 · 8

AuthorMessage
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13164
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1926472 - Posted: 25 Mar 2018, 1:29:53 UTC - in response to Message 1926467.  

Could have been online earlier in the day but I was waiting on the Fedex delivery for the monitor that Fedex said was "out for delivery" Finally figured they meant yes, it is on the truck . . . for Monday delivery.

Oh well, gave me chance to really stress the system for hours with Prime95 and memory stressapptest to be doubly sure the system was stable and wouldn't create errors.

Also allowed me to research why sensors-detect showed my sensor chip, the module was loaded but nothing was ever reported. Only the default coretemp module was showing anything. Finally figured out the sensor chip was at an address that Linux complained about having possible ACPI conflicts. Found the smoking gun in the dmesg output.

Juan, you have a X99 board and probably the same monitoring chip. Do you have the system monitoring chip reporting temps, fans and voltages? If not I can tell you how to get it fixed and reporting the sensor outputs that you see in the BIOS and Windows.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1926472 · Report as offensive
juan BFP Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 16 Mar 07
Posts: 9786
Credit: 572,710,851
RAC: 3,799
Panama
Message 1926475 - Posted: 25 Mar 2018, 1:34:51 UTC - in response to Message 1926472.  

Juan, you have a X99 board and probably the same monitoring chip. Do you have the system monitoring chip reporting temps, fans and voltages? If not I can tell you how to get it fixed and reporting the sensor outputs that you see in the BIOS and Windows.

Sure, but another day, with a lot less alcohol in my system. Nice it's working..
ID: 1926475 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13164
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1926480 - Posted: 25 Mar 2018, 1:59:44 UTC - in response to Message 1926475.  

Sure thing. It is an easy fix with just a couple words added to the grub file. Of course you would have needed to install the lm-sensors package in the first place and run the sensors-detect command to get the modules loaded. Once that is done, a simple edit to the grub linux command line and !Voila!, temps, fan speeds and voltages all appear in your favorite system monitor like Psensor or GKrellm.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1926480 · Report as offensive
Al Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Avatar

Send message
Joined: 3 Apr 99
Posts: 1682
Credit: 477,343,364
RAC: 482
United States
Message 1926488 - Posted: 25 Mar 2018, 3:00:51 UTC

Keith, I have an EVGA 1600 T2 powering my system with that motherboard, and with everything plugged in, I had just enough connectors to do the job. But it was close! Though only two dual 8 pin video cards are installed. That PSU also had the one to fire up that connector next to the 1st PCI-E slot, which I guess is nice, but I am having a weird issue that ASUS tech support is trying to help me out on, and would like me to take some pics of and sent them in to them.

Initially I had installed both 1080 Ti's in it, along with the Intel RAID card (more on that sad issue in a minute) and then loaded up Windows 7 Ultimate. As it was the original version of Ultimate, I had to install SP 1, then the rollup SP they released a year+ ago, and then started the process of loading up stuff on it. 1st off was the software for the RAID card, but that went off like a wet fart, so I spent a bit of time diagnosing that one, and came across a couple strange issues with the motherboard.

The main one was, in the advanced section of the BIOS, you can set your PCI-E slots to different speeds/specs - 1,2,3 and auto. Above that setting it lists what the slot is running at - x16, x8, etc. Well, on the bottom slot (I believe #6 at the time) it showed the 2nd video card installed at x16, and the RAID card as x8 a couple slots up, all good there. But for the 1080 in slot one, it said none installed.

So, I tried pulling out the lower 1080 and RAID card, and even pulled the battery out to completely nuke the CMOS and start over (on Asus's suggestion), and it still comes up with nothing installed in slot 1. Even though I am getting video out of it perfectly fine, Precision X controls it fine, no problems that I can find, but still, there is something wrong, it should show the card as something x16 is installed in that slot. Asus said to try pulling the battery again, and if that doesn't clear it up, to take some pics of some of the BIOS screens and send them in to their support dept and they will try and recreate the issue on their end. I'll probably try doing it again tomorrow and see if it makes any difference.

My concern is, that I have a basically brand new motherboard here, and if I end up sending it in because they can't identify the problem, the odds are very high that I will be getting a "Factory Certified Refurbished" (read: another persons problem board that they couldn't manage to repeat the issue on, so ship it out to the next poor fool) replacement board instead of a new one, and I had mentioned that concern to them when that possibility came up. They said if it came to that, they would send in a 'request' that I would like a new in the box board, and they said they couldn't promise anything.

If there is no firm solution to the problem, I am left with either just running it as is and hoping it doesn't cause problems down the road, especially when I am deep into working on a CAD file (save Often!), or sending it back to them and most likely getting a refurb'ed board, which I am not happy about. We'll have to see how it goes, though I do have to say that it's a pretty snappy system right now as configured, I guess I shouldn't beat myself up too badly about not spec'ing it exactly properly from the beginning, it'll be fine. :-)

The RAID card, well after having a couple discussions with Intel about the issues I had with seeing the card in Windows, (I had thought that the browser based software that is used to configure it also contained the hardware drivers as well) it turns out that this latest version of their RAID cards that were released last year no longer have any support for Windows 7. It has to be either 8.1, 10, or their server variants. Gaaa...

Yeah, I know, my fault for not researching it more closely, but as 7 still has almost 2 years left before the plug is pulled, well.. So anyways, I asked their support team if they could contact their driver team and see if they had possibly written a beta driver for Win 7 64 bit. I said that yes, I understood that it would be a use at my own risk kind of thing, etc etc, and they said they'd push the message up the food chain and get back to me next week. Not terribly hopeful with that, so..

I'm looking at possibly putting together a little system that I can put right next to it (but still using the main case to power and hold the 8 HDDs), using one of the 10GbE ports from the board, and going to a 10 GbE port on the new RAID one running right next to it. I spent a good chunk of yesterday searching for a mini or micro-ATX server board with at least one 10GbE port on it, and they are out there, but they sadly aren't cheap. If that RAID card wasn't so expensive, I'd just call it a learning experience and get one that is supported by 7, but at this point, I have to get it working. More on that later though.

ID: 1926488 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13164
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1926491 - Posted: 25 Mar 2018, 3:29:20 UTC

I saw that in the BIOS but left everything on Auto pretty much. I have had all four gpus in slots 1, 3, 5, and 7 showing in the BIOS as X16 speed. That is what is possible with the dual PLX chips and a 40 lane processor.

Your problem sounds eerily familiar with what Ruieke has been dealing with his ASUS TR board. He has had a Trouble Ticket in with them for a couple of months regarding the inability to set the the speed of one of the PCIe slots to X4 and have it stick. The BIOS says it makes the change and saves it only to not have the slot be active. This was when he was trying to run 6 of his ATI cards in the mobo.

You might want to compare notes with him and see if either of you can get better traction with ASUS.

I am not, and haven't been a fan of ASUS Tech support. They failed me with my Sabertooth R2.0 board and the X99-E -10G WS mobo made 5 trips through their RMA process. As you said, you just end up with someone else headache that wasn't resolved other than the merry-go-round of board swaps.

I detailed earlier in the thread about finding my X99 mobo incorrectly assembled with a missing PLX chip thermal pad material and that was unable to dissipate its heat and that it controlled the slots 5 and 7 that I was attempting to use in a stable BOINC crunching machine. I had to take it apart and fix the problem myself. There was no other way to fix it since the RMA actions were out of country and nothing could be done in the US ASUS tech support.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1926491 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13164
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1928019 - Posted: 4 Apr 2018, 20:44:41 UTC

Just one further update to the thread likely. I got the EVGA SuperNova 1600 T2 power supply too, Al. That gives me enough ports and connectors to hook everything up without all the add-on adapters I was having to use with the 1000W supply.

I tried to get one of the 1070's to work in the 4th slot on a powered PCIe slot riser but the system was flaky and hung several times just getting to the Desktop. I think the riser card added too much timing instability to the PCIe bus. So went back to the 1060 single fan card which fits directly in the slot and doesn't interfere with anything on the motherboard.

I have finalized on the cpu at 4.2Ghz. Think I am beginning to understand all the Intel BIOS parameters a bit. If the primary reason for this project was just to see what Intel looks like and the secondary reason was to add another cruncher to the farm, the experiment has been successful.

I see this cruncher as quickly becoming the fastest in the farm owing to the four gpus and the Linux special app in use. The only thing I might change in the future would be to replace the 1060 with one of the short form Zotac 1080's. But the 1060 on the special app is no slouch and that upgrade can be postponed far into the future.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1928019 · Report as offensive
juan BFP Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 16 Mar 07
Posts: 9786
Credit: 572,710,851
RAC: 3,799
Panama
Message 1928022 - Posted: 4 Apr 2018, 20:56:08 UTC - in response to Message 1928019.  
Last modified: 4 Apr 2018, 20:59:09 UTC

Obviusly we need more data, but for the few WU i was abble to find from your task list, crunched in the 1060 the crunching times rounds about 140-150 sec, not bad compared with the 90-100 secs of the 1080 or mine 1070 120-130secs.

If that is true your host runs stable at 4.2GHz it will pick the #2 slot of Seti top hosts very fast.
ID: 1928022 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13164
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1928054 - Posted: 5 Apr 2018, 1:08:32 UTC - in response to Message 1928022.  

Obviusly we need more data, but for the few WU i was abble to find from your task list, crunched in the 1060 the crunching times rounds about 140-150 sec, not bad compared with the 90-100 secs of the 1080 or mine 1070 120-130secs.

If that is true your host runs stable at 4.2GHz it will pick the #2 slot of Seti top hosts very fast.

Yes, I am not seeing that much of a performance degradement of the special app on the 1060 compared to the 1070Ti or the 1080. So I am not all that much in a hurry to upgrade to the Zotac Mini 1080. I'm not sure whether the bang for the buck is there for the purchase compared to the sunk cost of the 1060.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1928054 · Report as offensive
juan BFP Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 16 Mar 07
Posts: 9786
Credit: 572,710,851
RAC: 3,799
Panama
Message 1928056 - Posted: 5 Apr 2018, 1:14:16 UTC - in response to Message 1928054.  

Yes, I am not seeing that much of a performance degradement of the special app on the 1060 compared to the 1070Ti or the 1080.

I think i`m going to make some tests on my host, i have a 1060 here so will remove one of the 1070 and put the 1060 and leave running for some time and see the numbers. Let`s wait the weekend to try if i have time.
ID: 1928056 · Report as offensive
Previous · 1 . . . 5 · 6 · 7 · 8

Message boards : Number crunching : Come on admit it, you have an E5-2670 crunching on Boinc and/or Seti@Home. How is it doing?


 
©2024 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.