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 . . . 3 · 4 · 5 · 6 · 7 · 8 · Next

AuthorMessage
Al Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Avatar

Send message
Joined: 3 Apr 99
Posts: 1677
Credit: 426,082,543
RAC: 115,037
United States
Message 1924032 - Posted: 11 Mar 2018, 20:06:57 UTC

I actually took some pics of it when I just finished bolting the last pieces in, and then firing it up looking for any leaking Magic Smoke (none found, thank goodness!) but I'll have to figure out how to post them again, and where I had them hosted last time, since our site here doesn't host pics unfortunately.

ID: 1924032 · 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: 7585
Credit: 463,636,551
RAC: 11,485
Panama
Message 1924050 - Posted: 11 Mar 2018, 22:11:59 UTC
Last modified: 11 Mar 2018, 22:17:04 UTC

My main host runs in a TT X9 cube too. It's has a lot of space, the MB stays on the horizontal so no pain to the PCI-e slots with the GPU weight. Especially with a 4xGPU dessing of the X99 MB. A lot of room for the hybrids radiators & fans. I put all of them at the top of the case. They easily fix there and helps to create more flux for the air.
Just take care the case it's heavy I learn in the worst way when it fall on my finger. outch!!!
I'm sure you will going to enjoy to build your new toy on it.
ID: 1924050 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 8366
Credit: 732,882,626
RAC: 1,694,515
United States
Message 1924094 - Posted: 12 Mar 2018, 1:12:10 UTC - in response to Message 1924050.  

Juan would you say that your X99 motherboard runs hot? Meaning the VRM and PCH are hard to cool?
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1924094 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5012
Credit: 370,488,657
RAC: 414,293
United States
Message 1924099 - Posted: 12 Mar 2018, 1:19:03 UTC - in response to Message 1924094.  

Keith where are you monitoring that value at?
ID: 1924099 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 8366
Credit: 732,882,626
RAC: 1,694,515
United States
Message 1924102 - Posted: 12 Mar 2018, 1:33:03 UTC - in response to Message 1924099.  

A couple of different ways, my finger, the BIOS and my infrared temp gun. PCH runs at 70° C just idling in the BIOS. Temp hits 80° C. when stress testing the memory with all cores. And then the PC freezes up. Kept freezing just in the BIOS while I was looking around. Been reading about this board and I keep seeing posts that the PCH and VRM runs hot at just stock speed. The blame seems to fall on the extra PLX chip under the cover that most other X99 boards don't have. Been finding lots of other posts on more common X99 motherboards and the consensus is they run hot. I even found a post on my specific workstation board and the guy replaced the cheap ASUS thermal pad with some Fujipoly Extreme Thermal pad and dropped his temps by 15-20° C. I'm wondering if I should order some the same stuff and replace the stock pad under the PCH cover before I put everything together tomorrow and boot the system in its final configuration.

I finally realized the solution was to get my pedestal fan and blow over the front of the board while I was testing everything out on the table. No more freezes and reasonable temps. There is a big stock 200mm case fan in the front of the new chassis so I hope that is sufficient.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1924102 · 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: 7585
Credit: 463,636,551
RAC: 11,485
Panama
Message 1924145 - Posted: 12 Mar 2018, 9:53:56 UTC - in response to Message 1924094.  
Last modified: 12 Mar 2018, 9:58:30 UTC

Juan would you say that your X99 motherboard runs hot? Meaning the VRM and PCH are hard to cool?

No i never had problems with any of my X99 MB (i have one OC Formula from ASrock and and Gigabyte G1) about the hot temps, I use only the usual front big fan (the x9 cube comes with a 200 mm one) to keep the air flow mainly because the hot external temp i have here (normaly more than 30C). But i not OC anything. All the MB temps showed are below 58C (cpu and MB sensors) when crunching at 100%. I use a common Corsair WC with 2 fans (not remember the model but is one of the basic models) for the CPU.
ID: 1924145 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 8366
Credit: 732,882,626
RAC: 1,694,515
United States
Message 1924203 - Posted: 12 Mar 2018, 15:52:07 UTC - in response to Message 1924145.  

I was at box stock initial settings in the BIOS when I first powered the system and the PCH temps were too high. The stock locked cpu core frequency is 2.1GHz. The system CAN'T overclock the cpu. It is only a 85W TDP package. It is not running at any of the insane Intel i7 core clocks in the 4-5Ghz range or even close to what my Ryzens run at. This board just runs hot. It has the usual PCH chip under the cover PLUS another PLX chip to handle the extra PCIe lanes and the dual 10Gb network ports. That is a lot of extra heat to dissipate. I will stay with the stock chassis fans plus probably add a couple on the side nearest the graphics cards I intend to put into it to give them more air to breathe.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1924203 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 8366
Credit: 732,882,626
RAC: 1,694,515
United States
Message 1924629 - Posted: 14 Mar 2018, 22:36:19 UTC

I think there is something wrong with the X99 motherboard regarding the PCH temps. I put the system together with the stock Core X9 200mm fan in the front. I replaced the stock 120mm rear fan with a 140mm. I added a 140mm fan to the right side perforated panel side for more case airflow.

The system hung up in a few minutes of being powered on. Back into the BIOS to disable any fan controls and just have them run at full 12V and full rpm on all fans. No difference, the system hangs after a few minutes in the BIOS or the OS.

So I put back in the front drive cage and mounted a Corsair ML-140 fan onto the top of it with double sided tape to be at the same height as the motherboard and pointed it at the PCH cover and heatpipe sink thinking that was the easiest way to replicate my kitchen tabletop testing when I had a pedestal fan blowing over the front of the motherboard and the only way to keep the system from freezing.

I also removed the left side windowed panel to allow airflow from the pedestal fan blowing on the front and sides of the Ryzen 1800X and Xeon systems. I ordered another perforated side panel to replace the stock windowed side panel.

Problem solved if un-elegantly. So why does this motherboard require extraordinary cooling for just the PCH. The system is only using 225W at full load and only one graphics card installed. The system only runs at 2.3Ghz on all cores since it is a locked multiplier Xeon of only 85W TDP.

Been researching how to remove the PCH heat sinks and replacing the thermal pad material with something better. Anyone else have to do this with an X99 motherboard?
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1924629 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6530
Credit: 193,430,989
RAC: 7,363
United States
Message 1925456 - Posted: 20 Mar 2018, 2:48:27 UTC

Comparing the temps to my SC2600CP with two E5-2670's it may normal for that system.


SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the BP6/VP6 User Group today!
ID: 1925456 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 8366
Credit: 732,882,626
RAC: 1,694,515
United States
Message 1925459 - Posted: 20 Mar 2018, 4:30:47 UTC

Do you know which of those sensor labels would be the same as the PCH or chipset on my board?
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1925459 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6530
Credit: 193,430,989
RAC: 7,363
United States
Message 1925497 - Posted: 21 Mar 2018, 2:16:25 UTC

The BMC temp displays the same value as the PCH temp in other monitoring apps.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the BP6/VP6 User Group today!
ID: 1925497 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 8366
Credit: 732,882,626
RAC: 1,694,515
United States
Message 1925502 - Posted: 21 Mar 2018, 3:05:14 UTC - in response to Message 1925497.  

Thanks Hal. That is what the PCH temp reads in the BIOS when I have the extra cooling blowing over the board. When I first powered the board up without any cooling, the PCH went to 80° C. when fully loaded with Prime95. Locked up in within a couple of minutes after first powering up and going straight to stress testing. As long as I have the extra fan blowing onto the chip and its heat sinks the system is stable.

Haven't had any issues at all with BOINC crunching with two gpus since bringing it online. Granted, that is with the lowest possible performance out of the chip and memory. I will replacing the Xeon with a 6850K next weekend or the following week and I will be pushing the system a lot harder. I am also going to replace the thermal pads on the PCH and dual PLX chips when I upgrade the cpu.

Will find out whether the motherboard can handle the extra horsepower and still be stable. I have good hope it will still prove stable as long as the extra airflow is cooling the chipset. I also will probably pull the heat sink off the VRM's at least to check it out since the board is coming apart anyway. Will see whether the same thermal pad material is usable at that location also. I have two thicknesses of 1mm and 1.5mm which I learned is necessary for the PLX chips and the PCH chipset.

This has been an interesting experiment and learning process for me. Never done this kind of surgery on a motherboard before. If in the end I can get the 6850K working and the memory working at full XMP rating, and still be stable, it will have proved a good decision to resurrect a $650 motherboard.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1925502 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6530
Credit: 193,430,989
RAC: 7,363
United States
Message 1925722 - Posted: 22 Mar 2018, 2:34:29 UTC

I've considered regooping the VRMs and PCH but they are bolted down from the underside side of the MB.
So I would have to completed remove it & that is more work that I have been wanting to do.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the BP6/VP6 User Group today!
ID: 1925722 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 8366
Credit: 732,882,626
RAC: 1,694,515
United States
Message 1925728 - Posted: 22 Mar 2018, 3:22:26 UTC - in response to Message 1925722.  

Yes, it will entail taking the motherboard completely out of the system and demounting it from the motherboard tray to remove the heat sinks. I am planning to do that when I change cpus.

Today was just to get the new gpus installed and see what's what. Not doing so well right now. System is unstable. Trying to figure out why.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1925728 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 8366
Credit: 732,882,626
RAC: 1,694,515
United States
Message 1925888 - Posted: 22 Mar 2018, 23:24:32 UTC

Thought it at first it was just a bonehead action on my part with respect to power delivery on the 4th card. After correcting that I continued testing and have now come to the conclusion that the system just does not like that last PCIe X16 slot occupied. Ran all night long with 3 cards with no issues but as soon as I put in that 4th card, the system crashes within 15 minutes. System log shows the card dropping from the PCIe bus and the accumulation of errors eventually trips up the OS and it hangs. So until I find another short card to try or maybe a PCIe X16 bus riser to try a full length card, the system will comprise the cpu and three gpus.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1925888 · 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: 7585
Credit: 463,636,551
RAC: 11,485
Panama
Message 1925891 - Posted: 22 Mar 2018, 23:39:18 UTC - in response to Message 1925888.  

Did you remember to connect the extra power cable for the PCI-e slots?
ID: 1925891 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 8366
Credit: 732,882,626
RAC: 1,694,515
United States
Message 1925892 - Posted: 23 Mar 2018, 0:01:19 UTC - in response to Message 1925891.  

No I don't have any extra cables. I actually had to use a peripheral cable and molex to PCIe 6 pin adapter to power up the 1060. The power supply only has 6 PCIe ports on it and each 1080 takes two cables. I do have two EATX 12V cables plugged into the motherboard but I don't have any more PCIe cables.

I will have to order another peripheral cable and scrounge another Moles to PCIe 6 pin adapter to make another PCIe cable to power the PCIe slots. Once I do that I will attempt the 1060 again.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1925892 · 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: 7585
Credit: 463,636,551
RAC: 11,485
Panama
Message 1925893 - Posted: 23 Mar 2018, 0:09:12 UTC - in response to Message 1925892.  
Last modified: 23 Mar 2018, 0:12:47 UTC

Is a Asus X99 WS? IIRC it has an extra 6 pins PCI-e Conector at the right side of the slot 1 (the first one).

All the X99 MB i know needs the extra power connected when you run with 4 GPU's . Some uses molex connectors like mine, some use PCI-e or other type but all has one to help with the power the slots.

Check that could be the source of your instability.
ID: 1925893 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 8366
Credit: 732,882,626
RAC: 1,694,515
United States
Message 1925894 - Posted: 23 Mar 2018, 0:21:47 UTC - in response to Message 1925893.  

No its a ASUS X99-E-10G WS motherboard. Same as the Asus X99 WS except with 6 fewer SATA ports and dual 10G LAN ports.

I remembered the molex adapters in the 970 packages finally. I did in fact have another peripheral cable in the power supply cables.

So I will once again attempt to get the 1060 card running stable. Thanks for the tip about the extra PCIe slot power cable on the motherboard being necessary for all X99 boards using 4 cards.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1925894 · 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: 7585
Credit: 463,636,551
RAC: 11,485
Panama
Message 1925907 - Posted: 23 Mar 2018, 1:20:31 UTC - in response to Message 1925894.  
Last modified: 23 Mar 2018, 1:23:43 UTC

Thanks for the tip about the extra PCIe slot power cable on the motherboard being necessary for all X99 boards using 4 cards.

That was a High controvelsial theme in the past. The PCI-e specs said each PCI-e could power up to 75W and that is true, but when you draw power for all 4 slots, the total rises to up to 300W (with spikes of even more) on some moments. In theory that is OK, but in the real work the connectors has looses, and the PCB cooper layer has troubles to handle this high power spikes. That's why they put the extra connector close to the PCI-e slots. That helps with the current flow and fix the problem. Hope that fix your problem. At least on my case if i not connect the extra power the system is unstable like you post.
ID: 1925907 · Report as offensive
Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · 8 · Next

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


 
©2019 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.