MSI 560Ti HAWK - Maybe an issue on memory/VRM temperatures?

Hi.
Yesterday, I have decided to spend my evening to play at World of Warcraft (WoW). After I have played 2 hours without any problem (only some spare black pixel over the snow in a WoW landscape - I think a geometry problem), I have checked if all the data posted by Afterburner (voltages and temperatures) were ok.
So, I went to desktop and I have opened the MSI Afterburner to see temperatures and data.

The data was:
Core Clock: 950 Mhz
Shader Clock: 1900 Mhz
Memory Clock: 1903 Mhz
Fan Speed: 60
GPU: 41 °C (min: 32°C - max: 58°C)
Memory: 91 °C (min: 35°C - max: 95°C)
VRM: 49 °C (min: 33°C - max: 74°C)

Here, the screenshot where I have move down the core clock and speed up the fan.

When I see them, I immediately closed WoW because the memory temperature (I thought), were too high.
After few minutes, the GPU and VRM temperature slow down to 35 °C but the memory one still remained stable to 73 °C. (Here the screenshot)
After that, I shut down the system and I go to sleep.
This morning, before come to work, I run the system and all temperatures were 35 °C.

Is all Ok, or, like I think, something is not correct? Any ideas?
It is possible that the memory rise up to 95 °C? What is the max temperature that memory could reach? What is the correct temperature?
The VRM temperature is correct? It is possible that GPU and VRM have correct temperature and memory no?

Someone could post his memory/VRM temperature in idle and load?

On an italian forum, I read that someone saw, on the memory temperature graph of AB, an istant rise from 50°C to 60°C and from 70°C to 90 °C. So, I ask: the temp read by AB are ok or there could be some issues?
How are your memory/VRM temperature?

This is my pc configuration:

Case: Cooler Master CM 690 II Advanced
Fan position:
Top: 2 x 14 cm FAN Outtake;
Bottom: 1x 14 cm FAN Intake;
Front: 1 x 14 cm FAN Intake;
Rear: 1 x 12 cm FAN Outtake;
Side panel left: 1 x 14 cm FAN Intake;
Side panel right: 1 x 8 cm FAN Intake;
Hd case: 1 x 12 cm FAN

Mainboard: ASUS P8Z68-V PRO BIOS rev. 0606
CPU: Intel i7 2600K (original clock) + Thermalright Silver Arrow (2 x 14 cm FAN)
OS: Windows 7 Ultimate x64
RAM: Geil 16 GB (4x4 GB) + Arctic Rc Turbo Module
Audio: Creative X-fi XtremeMusic (PCI)
Video: MSI 560 GTX HAWK (original Clock)
PSU: LCPower - Arkangel 8850v2
HD: SATA6 HITACHI 1.5 TB + SATA6 WD Green 2.5 TB
DVD: Optiarc 7200S SATA

Driver Nvidia: 280.26
MSI Afterburner 2.2.0 beta 5

 

You should check if the max temp is a peak value or a stable one. If it's only peak one, you dont have to worry so much.

 

No, it is stable. As you can see from here, it is stable and it continues to rise up to 95 °C (after i close the application and I don't know if it could go higher).

 

Memory is waay to hot.

 

This evening I play newly at Wow.
I see a particular behaviour: the memory temperature proceds by steps an confirm what I found on the italian forum.
See the graph: in few second, the temperature passed from 58°C to 68°C. Is it normal?

 

I dont see the high ram temps with mine.
with auto fan and playing battlefield for a few hours
gpu 62c
mem 53c
vrm 58

its about 10c cooler then this test
9 mintutes with occt it does not change much if left longer




i concider your temps highly suspect..how can they run hotter then your vrm temps?

 

Another step: 77°C to 94°C in few seconds

Here another step: 77°C to 94°C in few seconds...
I really don't know how can i have these high temperature...
These steps are very strange: the temperature rise up by 10/20 °C in few seconds: it could be an error on the sensor? Or problem with the driver?

 

try a 275.xx driver just to see and make sure you dont have any other monitoring software going on. .
unlock the gpu voltage and monitoring in the settings just to see if that has anything to do with it?

i just added +80mv(1.67v) to the ram voltage with the fan up 100% like your run.
temps are even lower then auto fan??rams just dont run that hot




i'll test lastest driver for myself too.

edit
ok I tested the 280.33 with settings same as yours no mon./volt adding gpu uli. had cpuz,core temp,real temp, eleet,nvi and gpuz all open at the same time still same temps..

 

After 7 hours, during I have sleep, my pc stood in idle mode.
When I woke up (1 hour ago), I open AB, and this is the results. s you can see, the memory temperature is "blocked" at high level but I'm sure that those aren't the real temp of the memory compart. It's impossible that after 7 hours of idle (not stand-by), the temperature don't slow down.

I really think that remains only two options:
- the temperature sensor is break otherwise, it sends errata data to AB (someone kwon where is located the sensor on the VGA? I see that the thermal pads are bigger thant the memory chips: perhalps it could cover the sensor?!)
- something on my pc causes these errata values read by AB (a combination of driver?!)

I have to go to work. This evening I'll try to install 280.33 or 275.xx. Now, I'm running 280.26.

 

The screenshot shows the temperature of Memory is about 70c, I am not sure
the meaning of "the memory temperature is "blocked" at high level"

 

I'll search to explain.
After 7 hours during I didn't use the pc, the memory temperature still remained at 70°C, that I consider high level because there is no task that in 7 hours had used the VGA.

I have tried to reboot the system. After that, the temperature read by AB still remain the same but, if I shutdown the pc and then I turn on it (so, I don't make a reboot), the temperature read by AB is 35 °C.

What it could be?
I think that the memory thermal sensor is the problem. Could it be broken? Or some kind of "issue" between AB and nvidia driver?

 

i tested a few drivers w7 no issues.
do you have smart doctor installed?

 

Thanks for for help!
No, I haven't smartdoctor. Do you refer to Asus utility?
Is there a tool or a tecnique to check the sensor functionality?

 

well i'm not too smart or even know if this will help to find anything but..if you would use rivaturner and save your current report by going to customise>graphics subsystem dia. report.
check the same report categories i have in my report
here
http://www.mediafire.com/?cd1zpg85iodgvwu


we could see if there is any differances in
Display adapter PCI configuration space,graphic processor registers and your base address





BTW
did you try to take out your sound card for testing??

 

Maybe its a physical problem. Perhaps manufacturer didnt put thermal pads on it? You should take it apart and check

 

Sorry, but yesterday I couldn't check anything. I'll do all the test this afternoon/evening when I come back home.
Regarding the thermal pads it seems that they are all in the right place: there is only one that cover the chipsets and overflow, covering a small part of the the board.
I think that it could be "the problem" only if it close in itself the memory thermal sensor, so block the airflow and the sensor always remains hot, but I don't know where the sensor is located, on the VGA.

@Unwinder: it is important that I do the dump in this order?

1. dump when it reaches the highest temperature
2. shutdown the system (not reboot)
3. dump when the system starts and it is in idle

 

Hi,

i have 2 of these cards in my case running in SLI mode. I have the same kind of memory temps under load, same thing when idle, GPU is cool but memory still stays hot. I asked MSI support what are the max temps that GPU and memory can stand. Seems the GPU can reach 110° C and memory 115°C.

I noticed that MSI Afterburner (2.2.0 beta 5) only speed up fans for high GPU temps and it does not care about memory temps. That was confirmed by asking the MSI support.

I think there is a kind of bug with memory temps checking. This monitoring does not exist in previous MSI official AB release (2.1.0). I can't believe memory still stays hot when idling (GPUs are lower than 40°C) . However, the support told me to RMA my cards but i won't. A few days ago, i spent lot of time optimizing my case's airflow to make sure that was not the problem. From the beginning, i never had artifacts, freezes or some kinds of windows crashes.

I'll wait for a newer release of MSI AB for maybe an improvement. If any of my cards crash, i'll RMA it.

 

Do the temps sensors deal with the card's BIOS to report temps ? If it does, do you think it could be a BIOS issue ?

I ran MSI live update and upgraded the BIOS of my two cards from v111 to v112. Nothing had changed. But well, mine are running in SLI and are not far from each other (just one pci slot between them), Moreover, i guess the Hawk design (putting most of the heat in the case) doesn't help.