Greetings forum members. I have a DX79SR, on the latest bios, and am running the latest version of Desktop Utilities. I am experiencing an issue or anomaly that you may be able to help me with.
Most of the time, the Desktop Utilities report normal temperatures, 30 to 34 degrees C, for the # 1, # 2 and # 3 voltage regulators as well as the memory temperature. Recently, Desktop Utilities has flagged temperatures on the # 1 VR and the memory at 124 degrees C. A quick reboot without cutting power restores these readings to the previous normal. There seems to be no indication of overheating other than what Desktop Utilities is reporting.
Any idea if these readings are in error or are, for some reason, these two temperatures are actually spiking above boiling? I'd appreciate your thoughts.
Please verify you have installed the latest version of the Intel® Desktop Utilities
On the following link, please choose the Operating System you are using and check if it is the correct version.
You may need to manually uninstall the older version, reboot the computer, and then install the desired version.
I have tried uninstalling and reinstalling the software IDU v 3.3.6.068 and still the 124 degree readings on "Memory Temperature" and "CPU VR Temperature 1" jump to 124 degrees C intermittently causing IDU to flash an error and the fans to speed up to maximum. I have reset to default bios settings and am using the latest bios v.0590 dated 2013.7.17. I know the memory temperatures are normal because I have the "ambient air temperature" thermocouple on the memory modules.
Any other recommendations? Thanks in advance!
We have a machine that is reproducing the problem, so we likely won't need any additional information from you. It's looking like the SIO's Hardware Monitoring and Fan Speed Control subsystem just suddenly goes into the weeds. I am guessing it's a BIOS issue but am not sure. I will let you know once we have root caused the issue...
Thanks Sylvia and Scott. I did install 3rd party software just to confirm. I can't find any temperature info coming from CPU-Z v 1.65 so I installed something called CPUID HWMonitor. I can confirm that what IDU is seeing is also seen by HWMonitor in that the DRAM and the VREG temperatures also go maximum high (124 C). Must be the hardware/BIOS issue Scott is chasing down.
Incidentally, the DX79SR got into another state this morning where the CPU VR Temperature 1 went to 1 degree C in IDU and CPUID HWMonitor could no longer detect VREG (VR temperature). A reboot restored everything.
Thanks folks! I'll stay tuned.
An update: we've got a fix for the issue that we are in the process of verifying. If all goes well, we should be releasing it in a couple of days. Stay tuned...
Here's what I can tell you so far...
Upon close inspection of the fan speed control driver in the BIOS, I found two problems. Either one of these could result in an incorrect restoration of the SIO's state when coming out of S3. The worst of the two only affects the 8 Series boards - and indeed explains the issues being reported against the KL and MC boards (we are working on integrating the fix into new BIOSs for these boards). The other could affect just about any 6 Series or later board - and it sure looked like it could explain the problem being seen on the X79 boards. Alas, what it has done is just slow the rate of occurrence; it took 4 days for it to reoccur but reoccur it has. We are looking for an explanation for this; it appears that the SIO's fan speed control system is suddenly being shut off, causing incorrect temperature readings and fans racing to power-on defaults. How or what is doing/causing this is still TBD...
You may see a BIOS release that includes the two fixes - and indeed these may stop the issue from occurring for some people - but there's still something there that we haven't root-caused as yet...
Sorry, all I can say is we're working on it...
Hi Scott. I hate to bother you again, but this is becoming an intolerable situation with the fans racing to maximum several times a day and needing a reboot each time. Is the problem limited to all DX79SR boards or just a percentage of them? I ask because I wonder if I should I RMA this board with some level of certainty that the problem won't exist in the replacement?
As there has been no formal release of a new bios update, is there a beta bios that I can try to see if it mitigates or eliminates the problem?
Thanks in advance for your recommendation.
Hang in there; we're working on it. From what we can see, it is not a single board issue (I have seen it occur myself testing with both DX79SI and DX79SR boards). An RMA will not get you a board that doesn't have the problem, so that path is not the solution right now. We have a test BIOS with fixes for two separate bugs that were found during our investigation, but these fixes, while necessary, do not appear to address this issue that we are looking at (but, as we speak, these fixes are being rolled into BIOSs for other boards).
I will let you know where we stand once we sort it out...
Update: This has been tracked to an interoperability issue with certain third-party monitoring applications. Fortunately, the problem is on our side, not theirs, so we can do something about it. The forthcoming release of Intel(R) Desktop Utilities will provide a fix for this issue. I will update when we are closer to having this update ready...
Update: A new release of Intel(R) Desktop Utilities, v3.2.7.084, is available and contains a fix for the interoperability issue. We've verified it works with Finalwire's AIDA64 application and the CPUID organization's HWMonitor and PC Wizard 2013 applications. We still have a sighting open regarding interoperability with the SpeedFan application. I am in contact with Alfredo Milani-Comparetti, the SpeedFan author, and we are looking into this issue together...