Intel Socket 1151 Motherboards with IPMI AND Support for iGPU


Recommended Posts

In case it is helpful, here are the instructions for flashing the BMC with the socflash tool.  This is what I had to do when I inadvertently wiped out my BMC on board installation because of the depressed UID button:

 

I created a DOS bootable flash drive with Rufus, downloaded the linked files, then ran the flash command as in the example.

Begin forwarded message:

On Wednesday, February 26, 2020, 6:42 PM, [email protected] <[email protected]> wrote:

To reflash the BMC firmware using a DOS-executable utility – socflash. Please follow the steps below.

1/ Prepare a DOS-bootable USB flash drive.

2/ Download and save the unzipped socflash files onto the DOS-bootable USB flash drive. https://www.dropbox.com/s/kcua2ihf882d2gx/socflash_v11800.zip?dl=0

3/ Download the latest BMC firmware 1.60.0 from the download page and save the unzipped file onto the DOS-bootable flash drive.

https://www.asrockrack.com/general/productdetail.asp?Model=E3C246D4U#Download

4/ Boot up the board using your DOS-bootable USB flash drive. At the DOS 😄 prompt, please type the following.

Command to reflash BMC FW (DOS based): 

-socflash.exe if={bmc_image_file.ima}

for example, image file name = 111.ima

socflash.exe if=111.ima

After the BMC firmware programming is completed, please power off the board, then remove the 24-pin ATX power connector from the board for about 10 seconds prior to powering up the board again. Please let me know the result. Thank you!

 

  

Edited by Hoopster
Link to comment

As some of you know, running BOINC with the Rosetta@home project was causing my system to hard lock as, I believe, the Turbo Boost speed (4.6-4.8 GHz on ALL cores) was generating more heat than my CPU cooler could dissipate.  Since I am height limited in the case to a 146mm CPU cooler, I asked Noctua about the NH-U9S cooler (125mm) specifically with the E-2288G and this was their reply (it contains some interesting information):

 

"Thank you very much for contacting us.

 

That depends on your workload. Applications that heavily rely on any AVX instruction will push your CPU to a TDP of more than 240W, which is quite frankly ridiculously high. High temperatures on Coffee Lake CPUs with applications that use AVX instructions are inevitable. Adding a second fan to the NH-U9S usually yields a 2-4 °C improvement.

 

Therefore we can only partially recommend the NH-U9S. If you are running applications that heavily push the CPU (AVX) you are going to be looking at temperatures within the high 80s.

 

We'd recommend testing the max. as well as avg TDP in W for your CPU while you are using your computer as usual (we recommend using HWInfo).

Lastly, I'd like to point out that TDP ratings on CPU coolers can't be used to accurately measure the cooling capability. This is because the results are heavily influenced by numerous factors such as ambient temperature, size of the CPU, size and thickness of the IHS, thermal compound, heat density, target temperature etc.

 

Unfortunately, there is no standardised testing methodology for TDP testing."

Edited by Hoopster
Link to comment

I have a Norco 4220 case with a Noctua NH-D9L on my E-2288G, 110 mm height (fits in a 3U chassis). Even running all-core BOINC I haven't seen it go over 82C (ambient temp of around 65F). My system with all-core BOINC, all cores 100%, and two drives spun up is drawing 204 watts, according to my UPS.

 

I'm really pleased with it and it was the easiest heatsink install I've done to date.

 

https://www.amazon.com/gp/product/B00QCEWTAW

Edited by Dase
Added ambient temp and watt info.
Link to comment
18 minutes ago, Dase said:

I have a Norco 4220 case with a Noctua NH-D9L on my E-2288G, 110 mm height (fits in a 3U chassis). Even running all-core BOINC I haven't seen it go over 82 (ambient temp of around 65 degrees F).

I'm going to give the Noctua NH-U9S a try.  It arrived today.  My ambient temps are higher and the case is in a well-ventilated cabinet.  My current cooler is blowing air through the radiator fins back down towards the MB and CPU socket.  Perhaps it is combination of environmental and heat buildup through poor case airflow causing my problem.

 

The highest temp I recall seeing was 83C; however, I had over a dozen CPU lockups requiring a hard reset. 

 

Since I disabled Turbo Boost, I have had zero lockups and BOINC has been running non-stop for six days.

Link to comment
On 4/1/2020 at 3:37 PM, kaiguy said:

Update on the high MB temp. ASRock suggested I update to the latest BMC firmware.

Today the reported MB temp is 86C!  Doubt the 1.60 BIOS is going to help with those reports as that is what my board came with. 

 

In the six weeks I have had the board installed, that's the second time I have seen what has to be an erroneous reading.  There is no way the actual MB temp is 86C.  The CPU temp is currently 53C and I am supposed to believe the MB is 33C hotter?

Link to comment
23 minutes ago, Hoopster said:

Today the reported MB temp is 86C!  Doubt the 1.60 BIOS is going to help with those reports as that is what my board came with. 

 

In the six weeks I have had the board installed, that's the second time I have seen what has to be an erroneous reading.  There is no way the actual MB temp is 86C.  The CPU temp is currently 53C and I am supposed to believe the MB is 33C hotter?

Yeah, I figured the BIOS wasn't going to fix it since you've been on it since day one. If you have some time, can you maybe let ASRock know? He wanted me to update him if I get it again, and perhaps having more users reporting this MB temp issue would be helpful to get a solution...

Link to comment
On 4/2/2020 at 9:12 AM, Hoopster said:

As some of you know, running BOINC with the Rosetta@home project was causing my system to hard lock as, I believe, the Turbo Boost speed (4.6-4.8 GHz on ALL cores) was generating more heat than my CPU cooler could dissipate.  Since I am height limited in the case to a 146mm CPU cooler, I asked Noctua about the NH-U9S cooler (125mm) specifically with the E-2288G and this was their reply (it contains some interesting information):

 

"Thank you very much for contacting us.

 

That depends on your workload. Applications that heavily rely on any AVX instruction will push your CPU to a TDP of more than 240W, which is quite frankly ridiculously high. High temperatures on Coffee Lake CPUs with applications that use AVX instructions are inevitable. Adding a second fan to the NH-U9S usually yields a 2-4 °C improvement.

 

Therefore we can only partially recommend the NH-U9S. If you are running applications that heavily push the CPU (AVX) you are going to be looking at temperatures within the high 80s.

 

We'd recommend testing the max. as well as avg TDP in W for your CPU while you are using your computer as usual (we recommend using HWInfo).

Lastly, I'd like to point out that TDP ratings on CPU coolers can't be used to accurately measure the cooling capability. This is because the results are heavily influenced by numerous factors such as ambient temperature, size of the CPU, size and thickness of the IHS, thermal compound, heat density, target temperature etc.

 

Unfortunately, there is no standardised testing methodology for TDP testing."

I'm also running the NH-U9S, which seems to be adequate for my intended usage.  I'm not currently taking part in BOINC or FOLDINGATHOME, which I assume would potentially expose the shortcoming of this cooler.  With that said, I found it interesting the Noctua didn't exactly recommend a cooler.  So what cooler is recommended for the E-2288G at full load?

Link to comment
21 minutes ago, jbear said:

I'm also running the NH-U9S, which seems to be adequate for my intended usage.  I'm not currently taking part in BOINC or FOLDINGATHOME, which I assume would potentially expose the shortcoming of this cooler.  With that said, I found it interesting the Noctua didn't exactly recommend a cooler.  So what cooler is recommended for the E-2288G at full load?

Probably the NH-D15 (160mm) or the NH-U14S (165mm) as they both are rated at providing the best Turbo/Overclocking headroom for the i9-9900K which is the desktop equivalent of the E-2288G.  Unfortunately, both are too tall for my case and 146mm cooler headroom.

Edited by Hoopster
Link to comment
22 hours ago, kaiguy said:

Yeah, I figured the BIOS wasn't going to fix it since you've been on it since day one. If you have some time, can you maybe let ASRock know? He wanted me to update him if I get it again, and perhaps having more users reporting this MB temp issue would be helpful to get a solution...

After a reboot, Linux 'sensors' command and the System Temp plugin are reporting an MB temp of 29.5C.  In IPMI, it is still reporting 85C even after the reboot.

 

Checking the logs, the high MB temp has occurred three times.  After the fist two a reboot reset it to a normal temp reading.  This time, it is still stuck on 85C.

 

I have reported this to ASRock.

Link to comment
5 minutes ago, kaiguy said:

Considering just about all of us with this board here have experienced this, I'm surprised ASRock wasn't aware of this already.

Here is a response from ASRock I just received:

 

"I have a same E3C246D4U on my bench running everyday, but not necessarily a heavy load all the time. I check its BMC event log on regular basis but I haven’t yet been able to see such event log. I can always report this incident to our BMC engineering team, but they will ask me if I am able to duplicate this issue first, which I haven’t yet been able to."

 

I told him at least three of us in these forums have seen the issue.  He asked if I wanted a board replacement to see if it was an isolated issue.  I told him I did not think it was an isolated issue given the number on this forum who have seen it.  Perhaps it is related to heavy sustained loads which I suggested to him might be the case.

Edited by Hoopster
Link to comment
8 minutes ago, Burizado said:

I would say I don't have heavy sustained loads on my server.  My CPU typically is lower than 5%.  MB temp still says 91C via IPMI plugin and 27.8C on unRAID dashboard.

Well, that blows that theory.  Come to think of it, first time I saw it was with a light load as well.

 

Perhaps it would be helpful to report this to [email protected]

Link to comment

My Unraid server locked up last night, it was extracting a large multi file RAR, as far as I can tell anyway, this is not an unusual task, and have done this many times with the new setup, with no issues.  Not sure what to think of it, I suppose I will just monitor my server.  I'm running 6.9 BETA 1.

 

Outside of the BIOS settings relating to Onboard VGA for Plex transcoding, did you guys massage any of the other settings?  I looked through them based on my experience, and nothing jumped out at me, virtualization stuff was enabled by default etc.

 

I do have room for the larger cooler, before I spring for that, I want to make sure it's necessary.  No distributed computing dockers enabled.  No real CPU intensive tasks.  I may transcode 2-3 Plex streams  on occasion, all being doing via HW in RAM, which I have confirmed is working correctly.

 

JB

 

 

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.