CN102662914A - Method for configuring heat sensor of microprocessor - Google Patents
Method for configuring heat sensor of microprocessor Download PDFInfo
- Publication number
- CN102662914A CN102662914A CN2012101245454A CN201210124545A CN102662914A CN 102662914 A CN102662914 A CN 102662914A CN 2012101245454 A CN2012101245454 A CN 2012101245454A CN 201210124545 A CN201210124545 A CN 201210124545A CN 102662914 A CN102662914 A CN 102662914A
- Authority
- CN
- China
- Prior art keywords
- microprocessor
- focus
- temperature
- cluster
- heat sensor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Radiation Pyrometers (AREA)
Abstract
The invention relates to a method for configuring a heat sensor of a microprocessor, and the method comprises the steps of running a plurality of operating modes on the microprocessor, calculating the power consumption of the microprocessor under different operating modes by utilizing a power consumption calculating unit, and calculating the temperature distribution under the power consumption by utilizing a temperature distribution calculating unit; implementing data processing for the temperature distribution obtained under different operating modes, and acquiring a hot-spot distribution stacking chart of each module of the microprocessor; setting an upper limit value for the hot-spot monitoring error, performing the optimized double clustering considering singular points for data points in the hot-spot stacking chart according to the upper limit value, and obtaining a clustered result of the hot spot; and configuring one heat sensor for each clustering according to the clustered result, placing the heat sensor on a mass center of all hot spots contained by the clustering, namely, an average point of weight of the temperature value, and ensuring that the temperature of all hot spots are monitored within the set maximal error. Compared with the prior art, the method has advantages of high measurement precision and low cost.
Description
Technical field
The present invention relates to a kind of sensor configuration method, especially relate to a kind of microprocessor heat sensor configuration method.
Background technology
In the high performance circuit, owing to developing rapidly of miniaturization and LSI technology, the power density of microprocessor becomes increasingly high in modern times.These power attenuations change into heat rises chip temperature, causes reliability decrease, electricity leakage power dissipation increase, the cooling cost of microprocessor to rise.In order better to understand microprocessor ruuning situation, a lot of microprocessor manufacturers all adopt the thermal sensor reading of implanting on the sheet to assess the temperature conditions of the current operation of chip in the heat monitoring.Intel pentium 4 processors for example, it just has been equipped with thermal sensor, just can cause warning when temperature surpasses when setting working temperature, and processor is obtaining just reducing power consumption through closing clock temporarily after these report to the police.Therefore, how to distribute thermal sensor quantity and implantation position to become microprocessor heat management emphasis of design.
Through the prior art literature search is found; Ryan Cochran etc. proposed in 2009 a kind of utilize in the signal Processing spectrum technology with equally distributed sensor sample to dsc data carry out reconstruct; Recover 16 nuclear microprocessor heat distributions to reach purpose (Spectral Techniques for High-Resolution Thermal Characterization with Limited Sensor Data) to whole microprocessor heat monitoring; The document has provided a kind of high-precision microprocessor heat method for supervising; But be limited by the equally distributed requirement of sensor, certain limitation is arranged.Mukherjee etc. are to each module focus monitoring problem of microprocessor; Microprocessor heat sensor assignment and placement strategy (Systematic temperature sensor allocation and placement for microprocessors) based on the k-means clustering algorithm have been proposed; The document has been cast aside the whole heat distribution of microprocessor; Be placed on focal point on the focus of each module of microprocessor, the feasibility solution be provided for emergency mechanism designs with heat management.
Retrieval through to prior art finds that though these methods can provide a kind of focus monitoring solution, under high-precision requirement, the number of sensors that is adopted is still more, is unfavorable for saving design cost and manufacturing cost.Especially to the microprocessor of some high-performance high integration, because technological requirement, portion implants a large amount of sensors within it.How can guarantee that under the constant basically situation of precision, reducing number of sensors still is a problem demanding prompt solution.
Summary of the invention
The object of the invention is exactly to provide a kind of for the defective that overcomes above-mentioned prior art existence to guarantee effectively to reduce number of sensors, the microprocessor heat sensor configuration method that reduces cost under the high-precision requirement.
The object of the invention can be realized through following technical scheme:
A kind of microprocessor heat sensor configuration method, this method may further comprise the steps:
1) a plurality of operating modes of operation on microprocessor, the power consumption calculation unit calculates the power consumption of microprocessor under the different operating modes, and the Temperature Distribution computing unit calculates the Temperature Distribution of microprocessor under this power consumption;
2) temperature profile that obtains under the different operating modes is carried out data processing, obtain each module focus distribution stacking diagram of microprocessor;
3) set focus monitoring error higher limit, and the data point among the focus distribution stacking diagram is considered the dual cluster of optimization of singular point, obtain the cluster result of these focuses according to this higher limit;
4) according to above-mentioned cluster result, one piece of thermal sensor of each cluster configuration, thermal sensor is arranged on the barycenter place of contained all focuses of this cluster, i.e. and the average center of account temperature value weighting guarantees all hot(test)-spot temperature values of monitoring in setting maximum error.
Two-dimension temperature matrix when the Temperature Distribution of the microprocessor in the described step 1) is meant the microprocessor work state.
Described step 2) the focus distribution stacking diagram in is meant: according to the microprocessor architecture design Module Division, choose the focus of each module, all focuses under the different operating modes are distributed overlaps on the width of cloth Organization Chart again, and resulting focus distributes behind the rejecting repetition focus.
Focus error higher limit in the described step 3) is meant: in the design of microprocessor heat management, and the maximum difference of the thermal sensor sampling numerical value that is allowed and the focus actual temperature of monitoring.
Dual cluster in the described step 3) is meant the clustering algorithm that on spatial domain and Attribute domain, carries out simultaneously.
The dual cluster of optimization of the consideration singular point in the described step 3) is meant: dual clustering algorithm is chosen temperature value and is differed maximum data point than average temperature value when the initialization cluster centre.
Compared with prior art, the present invention has the following advantages:
1) to the heat distribution characteristics, adopt the resolution policy of self-organization bunch fusion, improved cluster efficient, reduce time cost, under the situation of many focuses, analysis speed is fast;
2) adopt the dual clustering algorithm of improvement of considering singular point, under equal accuracy required, the thermal sensor quantity of use was compared significantly with existing other schemes and is reduced, and cost is lower.
Description of drawings
Fig. 1 is a flow chart of steps of the present invention.
Embodiment
Below in conjunction with accompanying drawing embodiments of the invention are elaborated, present embodiment is to implement under the prerequisite invent technical scheme with this paper, provided detailed embodiment and concrete operating process, but protection scope of the present invention is not limited to following embodiment.
Embodiment 1
As shown in Figure 1, a kind of microprocessor heat sensor configuration method, this method may further comprise the steps:
The first step; Take Alpha EV6 microprocessor architecture design; The a plurality of operating modes of operation on microprocessor; Use SimpleScalar software emulation SPEC2000 (System Performance Evaluation Corporation) STP to simulate different operating modes, and pass through the power consumption of power consumption calculation unit computing module, the Temperature Distribution computing unit calculates the Temperature Distribution of microprocessor under this power consumption;
Above-mentioned power consumption calculation unit is a Wattch infrastructure software, and the Temperature Distribution computing unit is a Hotspot software.
In second step, the temperature profile that emulation under the different test procedures is obtained carries out focus merging and rejecting through MATLAB, obtains each module focus distribution stacking diagram of microprocessor, amounts to 20 focuses;
Said focus merges and rejecting is meant: be placed on the same Organization Chart getting hot spot data in all focus distribution plans, if there is hotspot location to overlap fully, so only keep one, other coincide points are rejected.
The 3rd step was set at 5% with maximum focus monitoring error, utilized the dual clustering algorithm of improvement of considering singular point to carry out the self-organization bunch Fusion of Clustering operation of all focuses, and its algorithm frame is following:
1) class center of initialization; This center be in all data points to be clustered with a maximum point of average difference, be the adjacent with it data point of reference scan with this center, whether the attributive distance that calculates these data points and center above threshold value; If attributive distance is less than threshold value; Then merge into one bunch, if attributive distance is greater than threshold value, then nonjoinder;
When 2) take place merging, immediately the data point set that newly is merged into as new class, any merging till not taking place in all adjacent data points around the scanning once more;
3) if one bunch can't merge with periphery again, then this bunch is exactly the complete cluster that we ask;
4) in remaining data point, circulate again 1,2,3 steps of execution, till all data points all belong to certain cluster.
The benefit of considering singular point is; In the process of cluster; Temperature value is compared often very difficult gathering with most of focus of the bigger focus of difference with average temperature value be one type, and coverlet Du Gu is upright, and causing need be its situation of distributing a sensor separately; Preferentially with it as cluster centre can be as much as possible periphery can type of permeating point distribute into this cluster, with the minimizing number of sensors.
Above-mentioned adjacent data point is meant: according to focus distribution superposition of data; Make the VORONOI figure of these focus dot matrix, if the VORONOI polygon accord with Q ueen criterion at certain two data points place, promptly there is public vertex in two polygons; So, this two data points then is judged as the consecutive number strong point.
Described attributive distance is meant: between two data points, and the Euclidean distance of non-space attribute.Suppose N d dimension strong point P
n={ g
n 1... g
n d, a
n 1... a
n m, 1≤n≤N; G wherein
n 1... g
n dBe the locus coordinate of this point, a
n 1... a
n mNon-space property value for this point.Any 2 P are then arranged
1, P
2Attributive distance be:
In the present embodiment, because the non-space property value has only temperature, so attributive distance is the point-to-point transmission temperature difference.
Said threshold value is meant: judging that whether two data points can gather is one type variable for one, is that the focus monitoring error is no more than and sets peaked key when guaranteeing to distribute sensor according to cluster result, so its threshold value itself is relevant with maximum heat point tolerance setting value.Suppose ε
MaxBe the maximum error ratio that engineering allows, n is the number of data points that cluster inside has contained, a
iIt is the non-space property value of all data points in the cluster.Then have:
The α here is an improvement factor, is used for balance sensor quantity and error, should choose flexibly according to different maximum error setting values, and present embodiment maximum error setting value is 5%, and the α value chooses 2.0.
In the 4th step, the number of clusters that finally obtains through above-mentioned steps is 4, and each cluster is distributed one piece of thermal sensor, promptly monitors all focuses according to 5% maximum error, and required thermal sensor quantity is 4.The position of sensor just is placed on the barycenter place of contained all focuses of this cluster, i.e. the average center of account temperature value weighting.
Embodiment 2
With reference to shown in Figure 1; The concrete steps of method are with embodiment 1; Difference is that setting maximum monitoring error respectively is 2%, 3% and 4%, and the number of sensors that adopts the inventive method and the resulting hot monitoring scheme of existing k-means clustering technique to adopt compares; The result is as shown in table 1, and the clear the inventive method of this illness that has not attacked the vital organs of the human body requires lower sensor quantity significantly to reduce at equal accuracy.When particularly accuracy requirement is higher, adopt the inventive method improved efficiency very obvious.(2% accuracy requirement is because near the limit of error, so promote not too obvious.)
Table 1
Claims (6)
1. microprocessor heat sensor configuration method is characterized in that this method may further comprise the steps:
1) a plurality of operating modes of operation on microprocessor, the power consumption calculation unit calculates the power consumption of microprocessor under the different operating modes, and the Temperature Distribution computing unit calculates the Temperature Distribution of microprocessor under this power consumption;
2) temperature profile that obtains under the different operating modes is carried out data processing, obtain each module focus distribution stacking diagram of microprocessor;
3) set focus monitoring error higher limit, and the data point among the focus distribution stacking diagram is considered the dual cluster of optimization of singular point, obtain the cluster result of these focuses according to this higher limit;
4) according to above-mentioned cluster result, one piece of thermal sensor of each cluster configuration, thermal sensor is arranged on the barycenter place of contained all focuses of this cluster, i.e. and the average center of account temperature value weighting guarantees all hot(test)-spot temperature values of monitoring in setting maximum error.
2. a kind of microprocessor heat sensor configuration method according to claim 1 is characterized in that the two-dimension temperature matrix the when Temperature Distribution of the microprocessor in the described step 1) is meant the microprocessor work state.
3. a kind of microprocessor heat sensor configuration method according to claim 1; It is characterized in that; Described step 2) the focus distribution stacking diagram in is meant: according to the microprocessor architecture design Module Division; Choose the focus of each module, all focuses under the different operating modes are distributed overlaps on the width of cloth Organization Chart again, resulting focus distribution plan behind the rejecting repetition focus.
4. a kind of microprocessor heat sensor configuration method according to claim 1; It is characterized in that; Focus error higher limit in the described step 3) is meant: in the design of microprocessor heat management, and the maximum difference of the thermal sensor sampling numerical value that is allowed and the focus actual temperature of monitoring.
5. a kind of microprocessor heat sensor configuration method according to claim 1 is characterized in that, the dual cluster in the described step 3) is meant the clustering algorithm that on spatial domain and Attribute domain, carries out simultaneously.
6. a kind of microprocessor heat sensor configuration method according to claim 1; It is characterized in that; The dual cluster of optimization of the consideration singular point in the described step 3) is meant: dual clustering algorithm is chosen temperature value and is differed maximum data point than average temperature value when the initialization cluster centre.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210124545.4A CN102662914B (en) | 2012-04-25 | 2012-04-25 | Method for configuring heat sensor of microprocessor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210124545.4A CN102662914B (en) | 2012-04-25 | 2012-04-25 | Method for configuring heat sensor of microprocessor |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102662914A true CN102662914A (en) | 2012-09-12 |
| CN102662914B CN102662914B (en) | 2015-01-28 |
Family
ID=46772410
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210124545.4A Expired - Fee Related CN102662914B (en) | 2012-04-25 | 2012-04-25 | Method for configuring heat sensor of microprocessor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102662914B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7263567B1 (en) * | 2000-09-25 | 2007-08-28 | Intel Corporation | Method and apparatus for lowering the die temperature of a microprocessor and maintaining the temperature below the die burn out |
| CN101093413A (en) * | 2006-06-21 | 2007-12-26 | 国际商业机器公司 | Heat regulation controlling method,system and processor used for testing real-time software |
-
2012
- 2012-04-25 CN CN201210124545.4A patent/CN102662914B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7263567B1 (en) * | 2000-09-25 | 2007-08-28 | Intel Corporation | Method and apparatus for lowering the die temperature of a microprocessor and maintaining the temperature below the die burn out |
| CN101093413A (en) * | 2006-06-21 | 2007-12-26 | 国际商业机器公司 | Heat regulation controlling method,system and processor used for testing real-time software |
Non-Patent Citations (3)
| Title |
|---|
| MUKHERJEE R等: "Systematic temperature sensor allocation and placement for microprocessors", 《PROCEEDINGS OF THE 43RD ANNUAL DESIGN AUTOMATION CONFERENCE》 * |
| SEDA OGRENCI MEMIK等: "Optimizing Thermal Sensor Allocation for Microprocessors", 《IEEE TRANSACTIONS ON COMPUTER-AIDED OF INTEGRATED CIRCUITS AND SYSTEMS》 * |
| XIN LI等: "Inverse Distance Weighting Method Based on a Dynamic Voronoi Diagram for Thermal Reconstruction with Limited Sensor Data on Multiprocessors", 《IEICE TRANS.ELECTRON.》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102662914B (en) | 2015-01-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9886326B2 (en) | Thermally-aware process scheduling | |
| CN102789246B (en) | Thermal protection method and device, and equipment with thermal protection function | |
| CN103020545B (en) | A kind of excess temperature protection method based on Loongson processor | |
| CN106050718A (en) | Intelligent fan control method and system | |
| Saranyasoontorn et al. | Design loads for wind turbines using the environmental contour method | |
| CN105787271B (en) | The adjustable output Interval evaluation method of thermal power plant unit based on big data analytical technology | |
| CN103062091B (en) | A kind of fan intelligent control method | |
| CN108960492A (en) | A kind of exhaust enthalpy of turbine prediction technique based on PSO-SVR soft-sensing model | |
| CN107085755A (en) | A kind of photovoltaic plant short term power Forecasting Methodology and system | |
| CN117075483B (en) | State monitoring method and device of liquid cooling energy storage system and storage medium | |
| CN102662914B (en) | Method for configuring heat sensor of microprocessor | |
| CN116341883B (en) | Resource coordination method and system for photovoltaic grid-connected switch | |
| CN112420228A (en) | Flow channel selection method and system for CHF (channel flow channel) of rod bundle channel under action of transient external force field | |
| CN108254670A (en) | For exchanging the health monitoring circuit structure of SoC at a high speed | |
| CN109740905A (en) | A kind of secondary probability of malfunction estimation method of polynary power grid based on Bayesian model | |
| CN102799707A (en) | Method and device for controlling sensors during dynamic heat management of micro processor | |
| US20170192491A1 (en) | Method for achieving low power consumption of three-dimensional measurement chip | |
| CN110532518B (en) | Air cooling contrast observation data interpolation method and system | |
| CN111030091B (en) | Method and system for determining installed electric capacity of distributed renewable energy | |
| Wang et al. | Comparison of different spatial resolution thermal infrared data in monitoring thermal plume from the Hongyanhe nuclear power plant | |
| CN102945506A (en) | Method for processing boundary contour image information for micro sitting of wind farm | |
| CN112702761A (en) | Method and system for detecting coverage hole of wireless sensor network | |
| CN103970239A (en) | On-chip dynamic intelligent temperature control system | |
| Liu et al. | A Hot Spot Identification Approach for Photovoltaic Module Based on Enhanced U-Net With Squeeze-and-Excitation and VGG19 | |
| CN202250996U (en) | Symmetric series-connection fan structure |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150128 Termination date: 20180425 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |