• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

K10 and 11 core temps

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Psykoikonov

Member
Joined
Dec 16, 2003
Location
ON, CA
After some debate, and my own interest into why my core temps idle well below ambient using an air cooler (on both a 1090T and an FX8120), I think I've found the answer. I continue to use the motherboard sensor and not the core sensor for the explanations given below.

The silicon and adhesives used in manufacturing these processors has a peak temperature rating of 97+ Celsius before any form of degradation will take place. The processor also has a thermal shut off safe guard in place that shuts the processor down at 90 Celsius.

The Cpu temperature is read form a sensor embedded within the socket of your motherboard causing about a 7-10 Celsius variance form the actual Cpu temperature, which may be what you are reading about on the net.

You can use an application called AMD overdrive, that will allow you to monitor your temperatures accurately.

As long as your core temperature has not exceeded the high side of the 60 degree mark for extended periods of time you should be ok. 62 degrees holds a generous safety net to begin with.

I hope I was able to answer your questions, If you have any more inquiries don't hesitate to contact us.


Thank You

Alex Cromwell
Senior Technology Director
Advanced Micro Devices
Fort Collins, Colorado
2950 East Harmony Road
Suite 300
Fort Collins, CO
80528-9558

and this taken from the K11 white paper
2.10.1 The Tctl Temperature Scale
Tctl is the processor temperature control value, used by the platform to control its cooling systems. Tctl is
accessible through SB-TSI and F3xA4[CurTmp]. Tctl is a non-physical temperature on an arbitrary scale measured
in degrees. It does not represent an actual physical temperature like die or case temperature. Instead, it
specifies the processor temperature relative to the point at which the system must supply the maximum cooling
for the processor’s specified maximum case temperature and maximum thermal power dissipation. It is defined
as follows for all parts:
• For Tctl = 0 to Tctl_max - 0.125: the temperature of the part is [Tctl_max - Tctl] degrees under the temperature
for which maximum cooling is required.
• For Tctl = Tctl_max to 255.875: the temperature of the part is [Tctl - Tctl_max] degrees over the worst-case
expected temperature under normal conditions. The processor may take corrective actions that affects performance
or operation as a result, such as invoking HTC or THERMTRIP_L.
2.10.2 Thermal Diode
The thermal diode is a diode connected to the THERMDA and THERMDC pins used for thermal measurements.
External devices use measurements from the thermal diode measurements to calculate temperature during
operation and test. These measurements are required to be adjusted as specified by F3xE4[DiodeOffset].
This diode offset supports temperature sensors using two sourcing currents only. Other sourcing current implementations
are not compatible with the diode offset and are not supported. A correction to the offset may be
required for temperature sensors using other current sourcing methods. Contact the temperature sensor vendor
to determine whether an offset correction is needed.

Most important is this
Tctl is a non-physical temperature on an arbitrary scale measured
in degrees. It does not represent an actual physical temperature like die or case temperature. Instead, it
specifies the processor temperature relative to the point at which the system must supply the maximum cooling
for the processor’s specified maximum case temperature and maximum thermal power dissipation.
 
Ah right..... so they are talking about the Cores when they say 62c, currently my cores max at 43c P-95 stressed, not that i will now be clocking another 500Mhz on it, its just good to know my CPU is running nice and cool with more OC room, and that's with my crappy CPU cooler.
 
I had one 965BE C3 whose cores read very close to ambient temps when it was idle, I had another 965BE C3 whose cores read ~15C below ambient...the core temp value as far as I'm concerned is completely arbitrary. The only thing I know for sure is that the motherboard sensor and the core temp go up the same amount when the cores are loaded on the two motherboards I've used since having a processor with core temp so far off.
 
IMO that's a bunch of mumbo jumbo nonsense. It is not physically possible to have core temps or CPU socket temps idle at a temp below ambient on air. If either core temps or CPU socket temps are being reported at below ambient on air then either the senors or the reporting software is miscalibrated.
 
Yeah, my own Core temps have reported below case temp in my old case, i know that's impossible.

But i will take it that 62c is the top Core temp tolerance, they must be using the same sensors to come up with that data.
 
Last edited:
Take it with a grain of salt if you prefer, but I have 2 processors in my current possession (1090T and FX8120) and one I sold (965BE C3) that using high end air cooling (Scythe Mugen II and Silver Arrow) report well below ambient core temps (10C-15C) at idle. The motherboard sensors that I've have used are far closer to the true temp given the fact that I know the ambient temp.
 
I always figure because the CPU temp is from the socket its reading from the side of the CPU, that bit is not being cooled, IE the underside.

I would think that will be hotter then the inside of the CPU which is where the core sensors are, and is the bit that's being cooled via the heat-spreader to the CPU cooler.

So for the CPU DIE and silicone i tend to trust the core temps, for the motherboard in the socket aria, Plastics ecte... i go on the CPU temp
 
It is no secret that with high end air cooling the core temps will typically be lower than CPU temps (for some obvious reasons) but I just can't buy the idea that the core temps would be colder than the environment that surrounds them when air cooling is used. It is also no secret that certain families of CPUs have core temp sensors that are miscalibrated significantly. One famous example the Athlon X2 64 CPUs with the Brisbane core from several years ago. Many also suspect that the Thubans are calibrated poorly and read too cool.
 
It is also no secret that certain families of CPUs have core temp sensors that are miscalibrated significantly.

This is what I'm trying to get at, and not only that but among the same family of CPUs they also differ, I had one 965BE C3 that read very close to correct temps (idle core temp was very close ambient temp) and another 965BE C3 that read ~15C below what the other said (15C below ambient at idle) in the exact same setting (room, case, motherboard, cooler, video card, etc, etc, etc.). However the motherboard temp read the same for both processors. On top of that the first 965BE (the one that read near correct temp) would fall apart (no stability regardless of more voltage) around the magical 55C core temp, and the second 965BE would fall apart at, you guessed it, around 40C core temp (40C+15=55C).
 
Last edited:
Well, I guess I misunderstood the point you were trying to make in the original post. I hope I didn't offend you. I thought you were trying to prove that you could cool something to a temperature lower than the temperature of what you were trying to cool it with.

Actually, now that you mention it this could explain why some people do not encounter instability when they cross the magical 55 C barrier.
 
No offence taken, it isn't easy to explain. I've known about for some time now but had nothing other than screenshots to back me up. I found the K11 white sheet that clearly explains that the core temp is an arbitrary number measured in degrees, so the only thing that can be taken from the core temp is the scale.
 
I think I understand that. A given temperature is like a pointer that is fixed because it is the reality part of the measuring process. What varies is where the scale is fixed in relation to the position of the pointer. In the manufacturing process it may be placed a little up or a little down from where it was placed the last time.
 
The K10 white paper says the same thing (no surprise).

2.10.1 The Tctl Temperature Scale
Tctl is the processor temperature control value, used by the platform to control cooling systems. Tctl is accessible
through SB-TSI and F3xA4[CurTmp]. Tctl is a non-physical temperature on an arbitrary scale measured in
degrees. It does not represent an actual physical temperature like die or case temperature. Instead, it specifies
the processor temperature relative to the point at which the system must supply the maximum cooling for the
processor’s specified maximum case temperature and maximum thermal power dissipation. It is defined as follows
for all parts:
• For Tctl = 0 to Tctl_max - 0.125: the temperature of the part is [Tctl_max - Tctl] degrees under the temperature
for which maximum cooling is required.
• For Tctl = Tctl_max to 255.875: the temperature of the part is [Tctl - Tctl_max] degrees over the worst-case
expected temperature under normal conditions. The processor may take corrective actions that affects performance
or operation as a result, such as invoking HTC or THERMTRIP_L.
2.10.2 Thermal Diode
The thermal diode is a diode connected to the THERMDA and THERMDC pins used for thermal measurements.
External devices use measurements from the thermal diode measurements to calculate temperature during
operation and test. These measurements are required to be adjusted as specified by F3xE4[DiodeOffset].
This diode offset supports temperature sensors using two sourcing currents only. Other sourcing current implementations
are not compatible with the diode offset and are not supported. A correction to the offset may be
required for temperature sensors using other current sourcing methods. Contact the temperature sensor vendor to determine whether an offset correction is needed. Feature support varies by package. See the Infrastructure Roadmap.

http://support.amd.com/us/Processor_TechDocs/31116.pdf
http://support.amd.com/us/Processor_TechDocs/41256.pdf
 
I think I understand that. A given temperature is like a pointer that is fixed because it is the reality part of the measuring process. What varies is where the scale is fixed in relation to the position of the pointer. In the manufacturing process it may be placed a little up or a little down from where it was placed the last time.

That sounds like the case, it may be up or down 15C from my experience (more than a little in my case).
 
Yeah, I've seen core temps and CPU temps vary by a much as 16 C. I suspect the real difference is more like about 5-6 degrees. When I see it vary much more than that I advise an offset correction.
 
Just to throw my experience with the amd chips in, its seems to me that the DTS sensors are not accurate @ idle at all. but the hotter the cores get the more accurate the DTS becomes. Using speedfan and core temp there is significant diffrence of 10-12 degrees@ idle, when the CPU is under full load the diffrence drops to withiin 2-3 degrees. in my experience.
 
X4 850 goes to 65c+ after 1 min of Prime95, which makes these "high"(well, high for an AMD CPU) safe-temp bars look like a joke.
Stock coolers are a no-go, the new pipe ones seem to keep them cool @ stock speeds, but old AM2+ ones that come with athlons just don't cut it.
 
Just to throw my experience with the amd chips in, its seems to me that the DTS sensors are not accurate @ idle at all. but the hotter the cores get the more accurate the DTS becomes. Using speedfan and core temp there is significant diffrence of 10-12 degrees@ idle, when the CPU is under full load the diffrence drops to withiin 2-3 degrees. in my experience.

That would make the statement in the white papers that at least the scale is right untrue as well. I don't think that is the case (could be wrong though). From the white paper I get that the idea is that the sensor is used for pwm cooling policy, so depending where the scale starts and where the max temp is set both can vary, but that the scale (degrees) would be the same from one processor to the next.
 
That would make the statement in the white papers that at least the scale is right untrue as well. I don't think that is the case (could be wrong though). From the white paper I get that the idea is that the sensor is used for pwm cooling policy, so depending where the scale starts and where the max temp is set both can vary, but that the scale (degrees) would be the same from one processor to the next.

I think there is only 1 DTS sensor for all your cores, not 1 per core.
 
Back