Techniques for accurately determining the temperature at various locations of an operating integrated circuit

Patent metadata

Field	Value
Publication number	US-2022026967-A1
Application number	US-202016934395-A
Country	US
Kind code	A1
Filing date	Jul 21, 2020
Priority date	Jul 21, 2020
Publication date	Jan 27, 2022
Grant date	—

Abstract

Integrated circuits (ICs)—depending on a current workload—may exceed thermal cooling budgets. As a result, ICs often implement thermal sensors to measure temperatures at junctions or hot spots along the IC. Due to a distance between the thermal sensors and the various junctions, a thermal offset may be added to the temperature readings from the thermal sensors to more accurately estimate the temperature at the junctions. To account for different workload distributions—e.g., asymmetric or symmetric—the systems and methods described herein may dynamically adjust the thermal offsets. As a result, the efficiency of the IC may be increased as thermal settings for the IC may take into account the ability of the thermal cooling budget to effectively cool the IC under a current operating condition—thereby reducing premature throttling back or shutting down of power to the IC.

First claim

1 . A method comprising:
computing an average temperature from outputs of a plurality of thermal sensors of an integrated circuit; determining a temperature difference between the average temperature and a temperature from an output of a thermal sensor of the plurality of thermal sensors; determining that the temperature difference is greater than a threshold value; and adjusting a first thermal offset value corresponding to at least the thermal sensor to a second thermal offset value based at least in part on the temperature difference being greater than the threshold value.

2 . The method of claim 1 , wherein the adjusting the first thermal offset to the second thermal offset value is executed incrementally over a plurality of time steps where temperature differences corresponding to at least one of the thermal sensor or another thermal sensor of the plurality of thermal sensors are greater than the threshold value.

3 . The method of claim 1 , further comprising:
computing an updated average temperature from additional outputs of the plurality of thermals sensors; determining that temperature differences from the additional outputs when compared to average temperatures of the plurality of thermal sensors are each less than the threshold value; and adjusting the second thermal offset value to the first thermal offset value based at least in part on the temperature differences being less than the threshold value.

4 . The method of claim 1 , wherein the second thermal offset value is greater than the first thermal offset value.

5 . The method of claim 1 , wherein the first thermal offset value and the second thermal offset value are used, at least in part, to determine final temperature values corresponding to one or more regions of the IC, and the final temperature values are compared against one or more temperature thresholds corresponding to at least one of throttling back power supplied to the IC, turning off power supplied to the IC, or increasing cooling for the IC.

6 . The method of claim 1 , further comprising:
determining a ratio between a current voltage supply of the IC to a maximum voltage supply of the IC; and determining that the ratio is above another threshold value, wherein the adjusting the first thermal offset value to the second thermal offset value is further based at least in part on the ratio being above the another threshold value.

7 . The method of claim 1 , further comprising:
determining a voltage to frequency ratio of at least a portion of the IC monitored by the temperature sensor; and determining that the voltage to frequency ratio is above another threshold value, wherein the adjusting the first thermal offset value to the second thermal offset value is further based at least in part on the voltage to frequency ratio being above the another threshold value.

8 . The method of claim 1 , wherein the adjusting the first thermal offset value to the second thermal offset value is for each of the plurality of thermal sensors.

9 . The method of claim 1 , wherein the adjusting the first thermal offset value to the second thermal offset value is for only the thermal sensor, and thermal sensors of the plurality of thermal sensors other than the thermal sensor have at least one of no adjustment to a respective thermal offset value or an individual adjustment to the thermal offset value separate from the adjusting the first thermal offset value to the second thermal offset value.

10 . The method of claim 1 , wherein values corresponding to the first thermal offset value and the second thermal offset are set during startup or initialization based at least in part on secure boot software.

11 . The method of claim 1 , wherein the IC corresponds to at least one of graphics processing unit (GPU), a central processing unit (CPU), a system on a chip (SOC), a sound chip, a network interface chip, a memory IC, a monolithic IC, a hybrid IC, a multi-chip IC, a digital IC, an analog IC, a mixed signal IC, a logic IC, an audio amplifier, an operational amplifier, a timer IC, a power management IC, an interface IC, or a combination thereof.

12 . The method of claim 1 , wherein values corresponding to the first thermal offset value and the second thermal offset value are determined using at least one of thermal simulations of a virtual representation of the IC or thermal observations of the IC.

13 . A method comprising:
computing, at each time step, an average temperature from outputs of a plurality of thermal sensors of an integrated circuit; computing, at each time step, a difference between the average temperature and a temperature corresponding to a thermal sensor of the plurality of thermal sensors; determining, at each time step, whether the difference is greater than a threshold value; and dynamically adjusting a thermal offset value of one or more of the plurality of thermal sensors to a first thermal offset value at a first subset of time steps where the difference is greater than the threshold value and to a second thermal value at a second subset of time steps where the difference is less than the threshold value.

14 . The method of claim 13 , wherein the first thermal offset value is greater than the second thermal offset value.

15 . The method of claim 13 , wherein the dynamically adjusting the thermal offset value is incremental from the thermal offset value toward one of the first thermal offset value or the second thermal offset value until the one of the first thermal offset value or the second thermal offset value are reached.

16 . The method of claim 13 , wherein thermal offset values are different for each of a first thermal sensor of the plurality of thermal sensors and a second thermal sensor of the plurality of thermal sensors.

17 . The method of claim 13 , wherein the dynamically adjusting the thermal offset value is applied universally to each of the plurality of thermal sensors.

18 . The method of claim 13 , wherein, at each time step, the dynamically adjusting the thermal offset is further based on at least one of a first ratio between current voltage supply to maximum voltage supply or a second ratio between voltage to frequency.

19 . A system comprising:
an integrated circuit including a first thermal sensor and a second thermal sensor; one or more processors; and one or more memory devices storing instructions that, when executed by the one or more processors, cause the one or more processors to execute operations comprising:
computing an average temperature between a first output of the first thermal sensor and a second output of the second thermal sensor;
determining computing a difference between the average temperature and a temperature from the first output;
determining that the difference is greater than a threshold value; and
adjusting a first thermal offset value corresponding to the first thermal sensor to a second thermal offset value based at least in part on the difference greater than the threshold value.

20 . The system of claim 19 , wherein the adjusting the first thermal offset value to the second value is executed incrementally over each consecutive time step where the difference is greater than the threshold value.

Assignees

• Nvidia Corp

Inventors

• Gu Jun
• Li Tao
• Plummer Chad
• Smith Brian Lawrence

Classifications

• Y02D10/00

  Energy efficient computing, e.g. low power processors, power management or thermal management · CPC title

• G06F11/3024 Primary

  where the computing system component is a central processing unit [CPU] · CPC title

• G06F11/3058

  Monitoring arrangements for monitoring environmental properties or parameters of the computing system or of the computing system component, e.g. monitoring of power, currents, temperature, humidity, position, vibrations (thermal management in cooling arrangements of a computing system G06F1/206) · CPC title

• G06F2200/201

  Cooling arrangements using cooling fluid · CPC title

• G06F1/324

  by lowering clock frequency · CPC title