- Aug 30, 2012
- 6,598
Thanks to Tom's Guide for this excellent article
Preface
Whether you overclock or not, the topic of processor temperatures can be very confusing. The purpose of this Guide is to provide an understanding of standards, specifications, thermal relationships and test methods so that temperatures can be uniformly tested and compared. This Guide supports Core i and Core 2 desktop processors running Windows Operating Systems.
Sections
1 - Introduction
2 - Ambient Temperature
3 - CPU Temperature
4 - Package Temperature
5 - Core Temperature
6 - Throttle Temperature
7 - Relative Temperatures
8 - Power and Temperature
9 - Overclocking and Voltage
10 - The TIM Problem
11 - Thermal Testing Tools
12 - Thermal Testing Basics
13 - Thermal Testing @ 100% Workload
14 - Thermal Testing @ Idle
15 - Improving Temperatures
16 - Summary
17 - References
Section 1 - Introduction
Intel desktop processors have a temperature for each Core, plus a temperature for the entire CPU, so a Quad Core has five temperatures. Heat originates at the transistor junctions within each Core where sensors measure Core temperatures. Depending on processor architecture, one of two different methods are used to measure CPU temperature.
Intel's Thermal Specification is "Tcase", which is CPU temperature, not Core temperature. Core temperature is 5C higher than CPU temperature due to differences in sensor proximity to the heat sources. For example, Tcase for the i5 4690K is 72C. Tcase + 5 makes the corresponding Core temperature 77C.
The relationship between Core temperature and CPU temperature is not in the Thermal Specifications; it's only found in a few engineering documents. In order to get a clear perspective of processor temperatures, it's important to understand the terminology and specifications.
Here’s a list of processors referenced according to microarchitecture:
6th Generation Core i, 14 nanometer
5th Generation Core i, 14 nanometer
4th Generation Core i, 22 nanometer
3rd Generation Core i, 22 nanometer
2nd Generation Core i, 32 nanometer
Previous Generation Core i, 32 nanometer
Previous Generation Core i, 45 nanometer
Legacy Core 2, 45 nanometer
Legacy Core 2, 65 nanometer
Use CPU-Z to identify your processor, then look up the specifications at Intel Product Information:
• CPU-Z - CPU-Z | Softwares | CPUID
• Intel Product Information - http://ark.intel.com
Section 2 - Ambient Temperature
Also called "room" temperature, this is the temperature measured at your computer's air intake. Standard Ambient temperature is 22C, which is normal room temperature. Ambient temperature is a reference value for Intel’s Thermal Specifications. Knowing your Ambient temperature is important because Ambient directly affects all computer temperatures. Use a trusted analog, digital or IR thermometer to measure Ambient temperature.
Here's the temperature conversions and a short scale:
Cx9/5+32=F ... or ... F-32/9x5=C ... or a change of 1C = a change of 1.8F
30.0C = 86.0F Hot
29.0C = 84.2F
28.0C = 82.4F
27.0C = 80.6F
26.0C = 78.8F Warm
25.0C = 77.0F
24.0C = 75.2F
23.0C = 73.4F
22.0C = 71.6F Norm ... or ... 22.2C = 72.0F
21.0C = 69.8F
20.0C = 68.0F
19.0C = 66.2F
18.0C = 64.4F Cool
When you power up your rig from a cold start, all components are at Ambient, so temperatures can only go up. With conventional air or liquid cooling, no temperatures can be less than or equal to Ambient.
As Ambient temperature increases, thermal headroom and overclocking potential decreases.
Section 3 - CPU Temperature
Also called "Tcase", this is the temperature shown in Intel's Thermal Specification. It's measured on the surface of the Integrated Heat Spreader (IHS) under tightly controlled laboratory conditions. For testing only, a groove is cut into the surface of the IHS where a "thermocouple" is embedded at the center, which accurately measures the temperature for the entire CPU. The stock cooler is then installed and the processor is tested at a steady 100% workload. One of two different methods are used to display “CPU” temperature in BIOS and in monitoring utilities.
Method 1: Legacy Core 2 (Socket 775) and Previous Generation Core i (Socket 1366) use a single Analog Thermal Diode centered under the Cores to substitute for a laboratory thermocouple. The Analog value is converted to Digital (A to D) by the motherboard's Super I/O (Input / Output) chip, then is calibrated to look-up tables coded into BIOS. Accuracy can vary greatly with BIOS updates. The monitoring utilities provided by motherboard manufacturers on the Driver DVD displays “CPU” temperature in Windows. For these processors, BIOS or CPU temperature may not be accurate.
Method 2: Previous Generation Core i and newer processors (Socket 115x and Socket 2011) no longer use an Analog Thermal Diode, but instead use the hottest Core as CPU temperature which is displayed in BIOS, and is defined as “Package” temperature (see Section 4). The monitoring utilities provided by motherboard manufacturers on the Driver DVD displays “CPU” temperature in Windows. For these processors, CPU temperature is the hottest Core, or Package temperature.
Regardless of the method used, CPU temperature in BIOS is higher than in Windows at idle, because BIOS boots the processor without power saving features to ensure that it will initialize under any conditions.
Section 4 - Package Temperature
Applies to: Previous Generation Core i and newer processors (Socket 115x and Socket 2011).
Package temperature is the hottest Core.
Package temperature is shown in a few software utilities such as Hardware Monitor - HWMONITOR | Softwares | CPUID - It can be affected by Intel's on-Die Integrated Graphics Processor Unit (IGPU).
Section 5 - Core Temperature
Also called "Tjunction", this is the temperature measured directly on the hot spots at the transistor junctions within each Core by individual Digital Thermal Sensors (DTS). Although sensors are factory calibrated by Intel, deviations between the highest and lowest Cores may be 10C. Sensors are more accurate at high temperatures to protect against thermal damage, so idle temperatures may not be accurate.
There's a 5C thermal gradient or "offset" between Core temperature and CPU temperature. This is shown on Figure 5 in the following Intel document - http://arxiv.org/ftp/arxiv/papers/0709/0709.1861.pdf
At Default / Auto BIOS settings (stock clock and Vcore) with 100% workload, Core temperature is 5C higher than the Tcase specification - ARK | Your Source for Intel® Product Specifications This means that whatever the Tcase specification is for your processor, add 5C to get the corresponding value for Core temperature.
Core temperature is the standard for thermal measurement.
Core temperatures respond instantly to changes in load.
Intel’s specification for DTS sensor response time is 256 milliseconds, or about 1/4th of a second. Since Windows has dozens of Processes and Services running in the background, it’s normal to see rapid and random Core temperature fluctuations, especially during the first few minutes after startup.
Here's the recommended operating range for Core temperature:
80C Hot (100% Load)
75C Warm
70C Warm (Heavy Load)
60C Norm
50C Norm (Medium Load)
40C Norm
30C Cool (Idle)
25C Cool
Core temperatures in the mid 70's are safe.
Your highest temperatures will occur when running test utilities. Temperatures are typically lower during real-world everyday workloads such as processor intensive applications or gaming.
Here’s a list of environment and hardware variables that affect Core temperature:
Ambient temperature
CPU cooler
Thermal Interface Material
Core voltage
Core speed
Memory
Computer location
Case design
Fans & ventilation
Cable management
GPU cooler
SLI / CrossFire
For more information see Section 15 - Improving Temperatures.
Section 6 - Throttle Temperature
Also called "Tj Max" (Tjunction Max), this is the Thermal Specification that defines the Core temperature at which the processor will Throttle (reduce clock speed) to protect against thermal damage. Although Intel processors are capable of operating above 90C, we also know that excessive heat kills electronics.
Sustained Core temperature greater than 80C is too hot for ultimate stability, performance and longevity.
Section 7 - Relative Temperatures
The relationships between CPU temperatures, Core temperatures and Throttle temperatures are shown below for several popular Quad Core processors, including load and idle Thermal Design Power (TDP). All values are based on Intel documentation.
-> Core i
6th Generation 14 nanometer: i7 6700K / i5 6600K (TDP 91W / Idle 2W)
Tcase (CPU temp) = 64C
Tjunction (Core temp) = 69C
Tj Max (Throttle temp) = 100C
5th Generation 14 nanometer: i7 5775C / i5 5675C (TDP 65W / Idle 2W)
Tcase (CPU temp) = 71C
Tjunction (Core temp) = 76C
Tj Max (Throttle temp) = 96C
4th Generation 22 nanometer: i7 4790K (TDP 88W / Idle 2W)
Tcase (CPU temp) = 74C
Tjunction (Core temp) = 79C
Tj Max (Throttle temp) = 100C
4th Generation 22 nanometer: i5 4690K (TDP 88W / Idle 2W)
4th Generation 22 nanometer: i7 4770K / i5 4670K (TDP 84W / Idle 2W)
Tcase (CPU temp) = 72C
Tjunction (Core temp) = 77C
Tj Max (Throttle temp) = 100C
3rd Generation 22 nanometer: i7 3770K / i5 3570K (TDP 77W / Idle 6W)
Tcase (CPU temp) = 67C
Tjunction (Core temp) = 72C
Tj Max (Throttle temp) = 105C
2nd Generation 32 nanometer: i7 2600K / i5 2500K (TDP 95W / Idle 8W)
Tcase (CPU temp) = 72C
Tjunction (Core temp) = 77C
Tj Max (Throttle temp) = 98C
Previous Generation 45 nanometer: i7 860 / i5 750 (TDP 95W / Idle 12W)
Tcase (CPU temp) = 72C
Tjunction (Core temp) = 77C
Tj Max (Throttle temp) = 100C
Previous Generation 45 nanometer: i7 920 D0 (TDP 130W / Idle 12W)
Tcase (CPU temp) = 67C
Tjunction (Core temp) = 72C
Tj Max (Throttle temp) = 100C
-> Core 2
Legacy 45 nanometer: Q9650 E0 (TDP 95W / Idle 16W)
Tcase (CPU temp) = 71C
Tjunction (Core temp) = 76C
Tj Max (Throttle temp) = 100C
Legacy 65 nanometer: Q6600 G0 (TDP 95W / Idle 24W)
Tcase (CPU temp) = 71C
Tjunction (Core temp) = 76C
Tj Max (Throttle temp) = 100C
Section 8 - Power and Temperature
The previous Sections have explained Intel’s Specifications, and how temperatures are measured and relate to one another. This Section will put the Specifications into perspective, and explain why Thermal Design Power, Tcase and Tj Max sometimes seem to conflict with recommended real-world Core temperatures.
Although Intel measures Tcase on the surface of the Integrated Heat Spreader (IHS), they also calculate the Tcase Specification based on their reference design stock coolers. There are several cooler models with different Thermal Design Power (TDP), which is expressed in Watts.
Certain TDP coolers are packaged with different TDP processors. For example, several Generations of Quad Core processors were packaged with a universal 95 Watt TDP cooler. When this 95 Watt cooler was matched with the 95 Watt i7 2600K, Tcase was calculated and measured at 72C. When the same 95 Watt cooler was packaged with the 77 Watt i7 3770K, Tcase was calculated and measured at only 67C (shown above in Section 7).
The i7 6700K and i5 6600K don’t include a stock cooler. Tcase is instead based on Intel’s new cooler which is sold separately: Intel’s Skylake Cooler - This is what Intel's first CPU cooler for Skylake looks like - Unlike the 95 Watt cooler packaged with earlier Quad Core processors, Intel’s new cooler is 130 Watts for the 91 Watt 6700K and 6600K, so Tcase was calculated and measured at a very low value of 64C (shown above in Section 7).
Tcase Specifications are determined by the stock cooler TDP. This is the primary reason why there’s so much variation in Tcase Specifications among processors of the same Core Count. For example, let’s compare two popular Quad Core CPU’s:
i7 6700K - 4 Cores / 8 Threads / 91 Watts TDP / Cooler 130 Watts TDP / Tcase = 64C.
i7 4790K - 4 Cores / 8 Threads / 88 Watts TDP / Cooler 95 Watts TDP / Tcase = 74C.
As you can see, although there's only a 3 Watt difference between the CPU's, there’s a 35 Watt difference in coolers which results in a 10C difference in Tcase specifications.
Additionally, processors without Hyperthreading (i5’s, Pentiums and Celerons) run cooler than their counterparts with Hyperthreading (i7’s and i3’s), even though they may have the same Tcase and TDP Specifications. The key word in the term “Thermal Design Power” (TDP) is Design. i5’s follow the i7 Design just as Pentiums and Celerons follow the i3 Design. The differences are Core Count, Cache, Instruction Sets and Hyperthreading, the latter of which has a pronounced effect on Core temperatures.
When sustained Core temperatures exceed 80C, some processors may become unstable. Even though Tj Max (Throttle temp) may be 100C for your CPU variant, it’s highly recommended that you avoid running your processor at Core temperatures that begin to approach Throttle temperature due to “Electromigration” (explained below in Section 9).
Core i 6th Generation processors have features such as Configurable TDP (cTDP) and Scenario Design Power (SDP) which may trigger throttling as low as 80C. See Sections 5.1.4 and 5.1.7 under “Thermal Management” in the following document: 6th Generation Intel Processor Datasheet - http://www.intel.com/content/dam/ww...sktop-6th-gen-core-family-datasheet-vol-1.pdf
Intel’s silicon fabrication has been very consistent for the past decade across dozens of variants. Processors with low Tcase specifications are just as thermally capable as the 88 Watt 4th Generation i7 4790K with a Tcase of 74C, which again is CPU temperature, not Core temperature; (Tcase + 5 = Core temperature).
So regardless of your processor’s microarchitecture, TDP, Tcase and Tj Max Specifications, BIOS settings, overclock, hardware configuration, CPU cooler, Ambient temperature, app / game / stress test software workloads or any other variables, here's the bottom line:
Core temperatures in the mid 70’s are safe. Sustained Core temperature greater than 80C is too hot for ultimate stability, performance and longevity.
Section 9 - Overclocking and Voltage
Overclocking is always limited by two factors; voltage and temperature. As Core speed (MHz) is increased above a level unique to each processor (silicon lottery), Core voltage (Vcore) must also be increased to maintain stability. This increases power consumption (Watts) which results in increased Core temperatures.
Overclocked processors using increased Vcore can run up to 50% above TDP. This is why high TDP air or liquid cooling is critical to keep Core temperatures under 80C. Overclocking should not be attempted with Vcore settings in “Auto” because BIOS will apply significantly more voltage than is necessary to maintain stability.
Even when using manual Vcore settings, excessive Vcore and temperatures may result in accelerated "Electromigration" - Google
This prematurely erodes the traces and junctions within the processor's layers and nano-circuits, which will eventually result in blue-screen crashes that become increasingly frequent over time. CPU's become more susceptible to Electromigration with each Die-shrink. However, Intel's advances in FinFET technology has improved the voltage tolerance of their 14 nanometer architecture.
Here’s a list of the maximum recommended Vcore settings:
-> Core i
6th Generation 14 nanometer ... 1.400 Vcore
5th Generation 14 nanometer ... 1.400 Vcore
4th Generation 22 nanometer ... 1.300 Vcore
3rd Generation 22 nanometer ... 1.300 Vcore
2nd Generation 32 nanometer ... 1.350 Vcore
Previous Generation 32 nanometer ... 1.350 Vcore
Previous Generation 45 nanometer ... 1.400 Vcore
-> Core 2
Legacy 45 nanometer ... 1.400 Vcore
Legacy 65 nanometer ... 1.500 Vcore
When tweaking your processor near it's highest overclock, keep in mind that for an increase of 100 MHz, a corresponding increase of about 50 millivolts (0.050) is needed to maintain stability. If 75 to 100 millivolts or more is needed for the next stable 100 MHz increase, it means your processor is overclocked beyond it's capability.
With high TDP air or liquid cooling you might reach the Vcore limit before 80C. With low-end cooling you’ll reach 80C before the Vcore limit. Regardless, whichever limit you reach first is where you should stop and declare victory. Testing is explained in Sections 11 through 14.
Remember to keep overclocking in perspective. For example, the difference between 4.4 GHz and 4.5 Ghz is less than 2.3%, which has no noticeable impact on overall system performance. It simply isn’t worth pushing your processor beyond recommended Core voltage and Core temperature limits just to squeeze out another 100 MHz.
Section 10 - The TIM Problem
Core i 3rd through 6th Generation processors are very sensitive to small increases in voltage and frequency. When overclocked, temperatures might exceed 80C, so high-end air or liquid cooling is critical. 3rd through 6th Generation processors are more difficult to cool than earlier processors for three reasons:
(1) The 3rd and 4th Generation 22 nanometer Die, and the 5th and 6th Generation 14 nanometer Die have significantly less surface area in contact with the underside of the Integrated Heat Spreader (IHS), than the larger 2nd Generation 32 nanometer Die.
(2) 3rd through 6th Generation processors have more transistors packed into a smaller Die than 2nd Generation processors.
(3) 3rd through 6th Generation processors use Thermal Interface Material (TIM) between the top of the Die and the underside of the IHS. Solder, which has superior thermal transfer characteristics, was instead used in 2nd Generation and earlier processors, and is used in Intel's "High End Desktop Processors" - Intel® High End Desktop Processors
http://i1275.photobucket.com/albums/y446/CompuTronix52/Package_zps43993989.jpg(Illustration from Intel Desktop 4th Gen Intel® Core™ Processors Datasheet, Vol. 1, Figure 24).
Since the bonding material which seals the perimeter of the IHS to the Substrate is slightly too thick, this tends to increase the space between the underside of the IHS and the Die, which can cause the TIM to compress unevenly. The effect of this manufacturing procedure is that many processors show a wide deviation between Core temperatures, or one Core which runs much hotter than it's neighbors.
This has encouraged some overclockers to "de-lid" or remove their processor's IHS, which basically involves thoroughly removing the bonding material, replacing only the TIM and then restoring the IHS. Typical results are significantly lower Core temperatures and less deviation between Cores. Here's an excellent YouTube - - that shows before, how-to, and after. Beware that de-lidding will void your warranty, and you can easily damage or destroy your processor.
Intel has addressed these thermal problems in their Haswell refresh. The Devil's Canyon processors have an improved IHS alloy and a new Polymer TIM. Although not as thermally efficient as solder, temperatures have been improved by several degrees.
Regardless, 4th Generation processors differ from their 3rd Generation counterparts in that they have a Fully Integrated Voltage Regulator (FIVR) on the Die, instead of on the motherboard, which increases their Thermal Design Power (TDP). Also, due to a 4.0 GHz clock speed, 4.4 GHz Turbo and increased Vcore, the 4th Generation 88 Watt Devil's Canyon i7 4790K runs hotter at 100% workload than any of it’s predecessors.
Note: 5th Generation 14 nanometer Broadwell processors also have a FIVR on the Die, but the TDP is much lower at only 65 Watts. 6th Generation 14 nanometer Skylake processors do not have Voltage Regulators on the Die. Even though TDP is 91 Watts, thermal behavior is similar to 3rd Generation Ivy Bridge 77 Watt processors.
Section 11 - Thermal Testing Tools
In order to properly test your temperatures, you'll need:
-> A trusted analog, digital or IR thermometer to measure Ambient temperature.
-> The following freeware utilities downloaded and installed -
• Core Temp - http://www.alcpu.com/CoreTemp
• CPU-Z - CPU-Z | Softwares | CPUID
• Prime95 v26.6 - Windows Downloads Center: Prime95 26.6
-> Optional; Install Real Temp (developed for Intel processors) to test your Core temperature sensors or monitor your temperatures - http://www.techpowerup.com/downloads/2089/real-temp-3-70
-> Optional; Install SpeedFan if you’d like to use the “Charts” to see your thermal signatures - http://www.almico.com/sfdownload.php
Section 12 - Thermal Testing Basics
We all remember science class where one of the guiding principles for conducting a controlled experiment, is that it's critical to follow the same procedure every time. This minimizes variables so results will be consistent and repeatable.
Since everyone tests their rigs with different hardware using X stress software at Y Ambient temperatures with Z measuring utilities resulting in CPU or Package or Core temperatures, it's impossible to compare apples to apples. This is why processor temperatures are so confusing.
There are only three relevant values; Ambient temperature, Core temperature at steady-state 100% workload, and Core temperature at dead idle. Applications, rendering, encoding and gaming are partial workloads with fluctuating temperatures which aren’t suitable for thermal testing or accurate temperature comparisons.
Sections 13 and 14 will explain how to properly test your rig at load and idle using standardized methods which minimize hardware, software and environmental variables. Follow the "Setup" in both Sections to replicate Intel's test conditions. Each 10 minute test will establish a valid thermal baseline.
Section 13 - Thermal Testing @ 100% Workload
Prime95 Version 26.6 Small FFT's is the standard for CPU thermal testing, because it's a steady-state 100% workload which runs Core 2 processors and Core i variants with Hyperthreading within 3% TDP at stock settings. This is the test that Real Temp uses to test sensors.
Core i 2nd through 6th Generation CPU's have AVX (Advanced Vector Extension) instruction sets. Recent versions of Prime95 such as 28.9 run AVX code on the Floating Point Unit (FPU) math coprocessor, which produces unrealistically high temperatures. The FPU test in the utility AIDA64 shows similar results.
Prime95 v26.6 produces temperatures on 3rd through 6th Generation processors more consistent with 2nd Generation, which also have AVX instructions, but do not suffer from thermal extremes due to having fewer transistors in a larger Die with greater surface area, and a soldered Integrated Heat Spreader.
Note: Keep in mind that we're thermal testing only. Stability testing is not within the scope of this Guide, which assumes your rig is stable. If you're overclocked, then a combination of stress tests, apps or games must be run to verify CPU stability.
If you’re overclocked and run AVX apps, you may need to reduce Vcore and clock speed and / or upgrade your cooling so that Core temperatures don’t exceed 80C. Asus RealBench runs a realistic AVX workload within 3% TDP at stock settings, however, it’s a cyclic workload for stability testing, which isn’t suitable for CPU thermal testing.
• Asus RealBench - http://rog.asus.com/rog-pro/realbench-v2-leaderboard/
Prime95's default test, Blend, is also a cyclic workload for testing memory stability, and Large FFT's combines CPU and memory tests. As such, Blend and Large FFT's both have cyclic workloads which aren’t suitable for CPU thermal testing.
Other stability tests such as Linpack and Intel Burn Test have cycles that peak at 110% workload, which again aren’t suitable for CPU thermal testing. The test utility OCCT runs elements of Linpack and Prime95, but will terminate the CPU tests at 85C.
The "Charts" in SpeedFan span 13 minutes, and show how each test creates different thermal signatures.
http://i1275.photobucket.com/albums/y446/CompuTronix52/SpeedFanTempGuideGraph_zpsd98effba.jpgShown above from left to right: Small FFT's, Blend, Linpack and Intel Burn Test.
Note the steady-state thermal signatures of Small FFT's, which allows accurate measurements of Core temperatures. A steady-state 100% workload is critical for thermal testing.
http://i1275.photobucket.com/albums/y446/CompuTronix52/SmallFFTsIntelETUAIDA64_zps2b0c9ff0.jpgShown above from left to right: Small FFT's, Intel Extreme Tuning Utility CPU Test, and AIDA64 CPU Test.
Intel Extreme Tuning Utility is also a cyclic workload. Although AIDA64's CPU test is steady-state, the workload is well below TDP, which is insufficient for thermal testing. All other AIDA64 CPU test combinations are cyclic workloads, which again aren't suitable for thermal testing.
Setup:
Testing should be performed with your computer clear of desk enclosures or items that block airflow. Covers should be removed and all fans and circulating pump (if equipped with liquid cooling) at 100% RPM, so temperatures can be tested under ideal conditions.
Testing close to 22C Ambient is preferred so as to provide normal thermal headroom, but is not required. During the summer climate, if adequate A/C isn’t available, then test late at night or early in the morning when Ambient is lowest.
Core temperatures rise and fall with Ambient temperature. When testing above or below 22C, it’s important to "normalize" test results to establish a valid thermal baseline. This minimizes variables so results will be consistent and repeatable.
Summer climates create above normal Ambient which decreases overclocking headroom. For example, if your measured Ambient is 4C above Standard, subtract 4C from your reported Core temperatures to normalize your test results.
Winter climates create below normal Ambient which increases overclocking headroom. For example, if your measured Ambient is 4C below Standard, add 4C to your reported Core temperatures to normalize your test results.
Core temperatures normalized to Standard Ambient are your baseline temperatures. Establishing a baseline is important because as Ambient changes, if you maintain your hardware configuration and BIOS settings, a baseline gives you a consistent point of reference. You can repeat the test whenever you like, to see if your rig is maintaining it’s thermal performance.
Test:
Run Prime95 v26.6 Small FFT's for 10 minutes, then use your thermometer to measure Ambient. Use Core Temp to measure your Core temperatures.
Results:
If reported Core temperatures exceed 80C, you should reduce Vcore and clock speed and / or improve cooling. Core temperatures in the mid 70's are safe.
Intel’s specification for Digital Thermal Sensor (DTS) accuracy is +/- 5C. This means deviations between the highest and lowest Cores may be 10C, so "Average" Core temperature is often more realistic.
On processors with more than 2 Cores, the inner Cores run warmer because they’re insulated by the outer Cores. Here’s the physical layout in 2nd, 3rd and 4th Generation Quad Core processors, and an example of how it typically affects Core temperatures:
IGPU = Not in use (PCIE graphics card in use)
Core #0 = 75C (insulated by IGPU and Core #1)
Core #1 = 78C (insulated by Core #0 and #2)
Core #2 = 76C (insulated by Core #1 and #3)
Core #3 = 71C (insulated by Core #2 only)
Core Average = 75C.
Note: When viewing your temperatures in Core Temp, values which reach 81C or higher will change from black to amber, which indicates caution.
Normalize your results to Standard Ambient and record the values for future reference.
Section 14 - Thermal Testing @ Idle
Look closely at the SpeedFan Charts above, where idle temperatures are shown between load temperatures. Note that some Cores have more "range" than others and idle lower. Sensors can be tested with Real Temp. Core temperature sensors are more accurate at high temperatures for Throttle protection, so idle temperatures may not be accurate.
If "Speedstep", also called Enhanced Intel Speedstep Technology (EIST), is disabled in BIOS, then depending on Vcore and clock speed, idle Power can be nearly 40 Watts, which will result in high idle temperatures, especially when combined with high Ambient temperature.
Setup:
In addition to using the previous Setup in Section 13, Speedstep and all "C" States must be enabled to achieve the lowest possible idle temperatures. Also, if Windows Power Options for "Balanced" or "Power saver" is not set correctly, then Speedstep will not work ... OR ... if Windows Power Options is set to "High performance", then Speedstep will not work because Minimum processor state can‘t be set.
To check this, click on Control Panel, Power Options, then to the right of the selected plan, click on Change plan setting. Next click on Change advanced power settings, then drag the scroll bar down. Click on + next to Processor power management, then click on + next to Minimum processor state. This Setting must be 5%. If it's not, then correct it and click Apply.
Restart into BIOS and confirm that you've saved your settings to a Profile. Next, change all settings to stock (Default / Auto) including SpeedStep, all C States and Vcore, then save and exit. Reboot into Windows and confirm that your rig is at dead idle; no programs running, and off line. No Folding or SETI or "tray-trash" running in the background, and less than 3% CPU Usage under the "Performance" tab in Windows Task Manager.
Use CPU-Z to confirm that Core Voltage and Core Speed has decreased as follows:
-> Core i
6th Generation 14 nanometer ... about 0.8 Volts @ 800 MHz
5th Generation 14 nanometer ... about 0.8 Volts @ 800 MHz
4th Generation 22 nanometer ... about 0.8 Volts @ 800 MHz
3rd Generation 22 nanometer ... about 0.9 Volts @ 1600 MHz
2nd Generation 32 nanometer ... about 1.0 Volts @ 1600 MHz
Previous Generation 32 nanometer ... about 1.0 Volts @ 1600 MHz
Previous Generation 45 nanometer ... about 1.0 Volts @ 1600 MHz
-> Core 2
Legacy 45 nanometer ... about 1.1 Volts @ 2000 MHz
Legacy 65 nanometer ... about 1.25 Volts @ 1600 MHz
Use Core Temp to confirm that Power has decreased as follows:
-> Core i
6th Generation 14 nanometer ... about 2 Watts
5th Generation 14 nanometer ... about 2 Watts
4th Generation 22 nanometer ... about 2 Watts
3rd Generation 22 nanometer ... about 6 Watts
2nd Generation 32 nanometer ... about 8 Watts
Previous Generation 32 nanometer ... about 8 Watts
Previous Generation 45 nanometer ... about 12 Watts
Note: Idle Volts and Watts may differ depending on BIOS versions and motherboard models. Power (Watts) isn't measured on Previous Generation Core i Socket 1366 variants and Legacy Core 2 processors, but for general reference, idle power for several popular CPU's is shown in Section 7 - Relative Temperatures.
Test:
Allow your rig to "settle" for 10 minutes, then use your thermometer to measure Ambient. Use Core Temp to measure your Core temperatures.
Results:
Core i 2nd through 6th Generation processors should idle at less than 8C above Ambient. This means at 22C Standard Ambient your Cores should idle just under 30C. Certain Previous Generation Core i variants and Legacy Core 2 processors may idle several degrees higher. Better cooling and lower idle power produce lower idle temperatures.
Normalize your results to Standard Ambient and record the values for future reference. When finished testing, restore your system to it's previous configuration.
Section 15 - Improving Temperatures
Whether your computer is a stock workstation or an overclocked gaming rig, achieving the lowest possible temperatures always depends on components, configuration and airflow. Here's a few thoughts:
• Intel coolers are barely adequate at stock. If you want to overclock then upgrade your cooler.
• Use Manual Vcore settings. Auto applies excess voltage which means more power and heat.
• Memory overclock and XMP Profiles can cause Core i processors to run a few degrees hotter.
• Axial flow graphics cards recirculate heat. Linear flow cards exhaust heat from the case.
Examples:
Axial - http://www.newegg.com/Product/Product.aspx?Item=N82E16814487248
Linear - http://www.newegg.com/Product/Product.aspx?Item=N82E16814487247
• Axial cards work well with a liquid cooled CPU. Linear cards work well with an air cooled CPU.
• SLI / CrossFire works best with Linear cards. Axial cards dump massive heat in your case.
• A hot case stresses hard drives, memory, chipsets, voltage regulators and power supply.
• High performance computers need unrestricted airflow in and out, so location is critical.
• Load temperatures that drop over a degree or two with case covers off means poor airflow.
• Good cable management creates good airflow. Use zip-ties, patience and attention to detail.
• Quality fans are important, but if you want a quiet computer then consider a fan controller.
• If your case just doesn't breathe well, then perhaps it's time to upgrade to one that does.
• If your rig runs 24/7, then hard drive and fan bearings are wearing, and dust is accumulating.
• Clean the dust out of your rig. Perform regular Planned Maintenance Inspections (PM's).
• Replace your TIM. Most Thermal Interface Material typically begins to fail after 2 years.
Thermal Interface Material (TIM):
Thermal Paste Comparison, Part One: Applying Grease And More - http://www.tomshardware.com/reviews/thermal-paste-heat-sink-heat-spreader,3600.html
Thermal Paste Comparison, Part Two: 39 Products Get Tested - http://www.tomshardware.com/reviews/thermal-paste-performance-benchmark,3616.html
The proper installation of Intel's stock cooler:
Intel Stock Cooler Installation Guide - http://www.tomshardware.com/forum/338655-28-intel-stock-cooler-installation-guide
Choosing an aftermarket cooler:
Air Cooling vs Water Cooling : Things You Need To Know - http://www.tomshardware.com/forum/id-2196038/air-cooling-water-cooling-things.html
Alternatives to the Hyper 212+/Evo for budget cooling - http://www.tomshardware.com/forum/id-2705157/alternatives-hyper-212-evo-budget-cooling.html
Section 16 - Summary
• Standard Ambient temperature is 22C.
• Ambient affects all computer temperatures.
• No temperatures can be less than or equal to Ambient.
• As Ambient increases, thermal headroom decreases.
• BIOS or CPU temperature may not be accurate.
• Package temperature is the hottest Core.
• Core temperature is the standard for thermal measurement.
• Core temperatures respond instantly to changes in load.
• 80C sustained Core temperature is too hot.
• Core temperatures in the mid 70's are safe.
• Excessive Vcore and temperatures accelerate electromigration.
• Prime95 v26.6 Small FFT's is the standard for thermal testing.
• Deviations between highest and lowest Cores may be 10C.
• Core temperature sensors are more accurate at high temperatures.
• Idle temperatures may not be accurate.
• Sensors can be tested with Real Temp.
Preface
Whether you overclock or not, the topic of processor temperatures can be very confusing. The purpose of this Guide is to provide an understanding of standards, specifications, thermal relationships and test methods so that temperatures can be uniformly tested and compared. This Guide supports Core i and Core 2 desktop processors running Windows Operating Systems.
Sections
1 - Introduction
2 - Ambient Temperature
3 - CPU Temperature
4 - Package Temperature
5 - Core Temperature
6 - Throttle Temperature
7 - Relative Temperatures
8 - Power and Temperature
9 - Overclocking and Voltage
10 - The TIM Problem
11 - Thermal Testing Tools
12 - Thermal Testing Basics
13 - Thermal Testing @ 100% Workload
14 - Thermal Testing @ Idle
15 - Improving Temperatures
16 - Summary
17 - References
Section 1 - Introduction
Intel desktop processors have a temperature for each Core, plus a temperature for the entire CPU, so a Quad Core has five temperatures. Heat originates at the transistor junctions within each Core where sensors measure Core temperatures. Depending on processor architecture, one of two different methods are used to measure CPU temperature.
Intel's Thermal Specification is "Tcase", which is CPU temperature, not Core temperature. Core temperature is 5C higher than CPU temperature due to differences in sensor proximity to the heat sources. For example, Tcase for the i5 4690K is 72C. Tcase + 5 makes the corresponding Core temperature 77C.
The relationship between Core temperature and CPU temperature is not in the Thermal Specifications; it's only found in a few engineering documents. In order to get a clear perspective of processor temperatures, it's important to understand the terminology and specifications.
Here’s a list of processors referenced according to microarchitecture:
6th Generation Core i, 14 nanometer
5th Generation Core i, 14 nanometer
4th Generation Core i, 22 nanometer
3rd Generation Core i, 22 nanometer
2nd Generation Core i, 32 nanometer
Previous Generation Core i, 32 nanometer
Previous Generation Core i, 45 nanometer
Legacy Core 2, 45 nanometer
Legacy Core 2, 65 nanometer
Use CPU-Z to identify your processor, then look up the specifications at Intel Product Information:
• CPU-Z - CPU-Z | Softwares | CPUID
• Intel Product Information - http://ark.intel.com
Section 2 - Ambient Temperature
Also called "room" temperature, this is the temperature measured at your computer's air intake. Standard Ambient temperature is 22C, which is normal room temperature. Ambient temperature is a reference value for Intel’s Thermal Specifications. Knowing your Ambient temperature is important because Ambient directly affects all computer temperatures. Use a trusted analog, digital or IR thermometer to measure Ambient temperature.
Here's the temperature conversions and a short scale:
Cx9/5+32=F ... or ... F-32/9x5=C ... or a change of 1C = a change of 1.8F
30.0C = 86.0F Hot
29.0C = 84.2F
28.0C = 82.4F
27.0C = 80.6F
26.0C = 78.8F Warm
25.0C = 77.0F
24.0C = 75.2F
23.0C = 73.4F
22.0C = 71.6F Norm ... or ... 22.2C = 72.0F
21.0C = 69.8F
20.0C = 68.0F
19.0C = 66.2F
18.0C = 64.4F Cool
When you power up your rig from a cold start, all components are at Ambient, so temperatures can only go up. With conventional air or liquid cooling, no temperatures can be less than or equal to Ambient.
As Ambient temperature increases, thermal headroom and overclocking potential decreases.
Section 3 - CPU Temperature
Also called "Tcase", this is the temperature shown in Intel's Thermal Specification. It's measured on the surface of the Integrated Heat Spreader (IHS) under tightly controlled laboratory conditions. For testing only, a groove is cut into the surface of the IHS where a "thermocouple" is embedded at the center, which accurately measures the temperature for the entire CPU. The stock cooler is then installed and the processor is tested at a steady 100% workload. One of two different methods are used to display “CPU” temperature in BIOS and in monitoring utilities.
Method 1: Legacy Core 2 (Socket 775) and Previous Generation Core i (Socket 1366) use a single Analog Thermal Diode centered under the Cores to substitute for a laboratory thermocouple. The Analog value is converted to Digital (A to D) by the motherboard's Super I/O (Input / Output) chip, then is calibrated to look-up tables coded into BIOS. Accuracy can vary greatly with BIOS updates. The monitoring utilities provided by motherboard manufacturers on the Driver DVD displays “CPU” temperature in Windows. For these processors, BIOS or CPU temperature may not be accurate.
Method 2: Previous Generation Core i and newer processors (Socket 115x and Socket 2011) no longer use an Analog Thermal Diode, but instead use the hottest Core as CPU temperature which is displayed in BIOS, and is defined as “Package” temperature (see Section 4). The monitoring utilities provided by motherboard manufacturers on the Driver DVD displays “CPU” temperature in Windows. For these processors, CPU temperature is the hottest Core, or Package temperature.
Regardless of the method used, CPU temperature in BIOS is higher than in Windows at idle, because BIOS boots the processor without power saving features to ensure that it will initialize under any conditions.
Section 4 - Package Temperature
Applies to: Previous Generation Core i and newer processors (Socket 115x and Socket 2011).
Package temperature is the hottest Core.
Package temperature is shown in a few software utilities such as Hardware Monitor - HWMONITOR | Softwares | CPUID - It can be affected by Intel's on-Die Integrated Graphics Processor Unit (IGPU).
Section 5 - Core Temperature
Also called "Tjunction", this is the temperature measured directly on the hot spots at the transistor junctions within each Core by individual Digital Thermal Sensors (DTS). Although sensors are factory calibrated by Intel, deviations between the highest and lowest Cores may be 10C. Sensors are more accurate at high temperatures to protect against thermal damage, so idle temperatures may not be accurate.
There's a 5C thermal gradient or "offset" between Core temperature and CPU temperature. This is shown on Figure 5 in the following Intel document - http://arxiv.org/ftp/arxiv/papers/0709/0709.1861.pdf
At Default / Auto BIOS settings (stock clock and Vcore) with 100% workload, Core temperature is 5C higher than the Tcase specification - ARK | Your Source for Intel® Product Specifications This means that whatever the Tcase specification is for your processor, add 5C to get the corresponding value for Core temperature.
Core temperature is the standard for thermal measurement.
Core temperatures respond instantly to changes in load.
Intel’s specification for DTS sensor response time is 256 milliseconds, or about 1/4th of a second. Since Windows has dozens of Processes and Services running in the background, it’s normal to see rapid and random Core temperature fluctuations, especially during the first few minutes after startup.
Here's the recommended operating range for Core temperature:
80C Hot (100% Load)
75C Warm
70C Warm (Heavy Load)
60C Norm
50C Norm (Medium Load)
40C Norm
30C Cool (Idle)
25C Cool
Core temperatures in the mid 70's are safe.
Your highest temperatures will occur when running test utilities. Temperatures are typically lower during real-world everyday workloads such as processor intensive applications or gaming.
Here’s a list of environment and hardware variables that affect Core temperature:
Ambient temperature
CPU cooler
Thermal Interface Material
Core voltage
Core speed
Memory
Computer location
Case design
Fans & ventilation
Cable management
GPU cooler
SLI / CrossFire
For more information see Section 15 - Improving Temperatures.
Section 6 - Throttle Temperature
Also called "Tj Max" (Tjunction Max), this is the Thermal Specification that defines the Core temperature at which the processor will Throttle (reduce clock speed) to protect against thermal damage. Although Intel processors are capable of operating above 90C, we also know that excessive heat kills electronics.
Sustained Core temperature greater than 80C is too hot for ultimate stability, performance and longevity.
Section 7 - Relative Temperatures
The relationships between CPU temperatures, Core temperatures and Throttle temperatures are shown below for several popular Quad Core processors, including load and idle Thermal Design Power (TDP). All values are based on Intel documentation.
-> Core i
6th Generation 14 nanometer: i7 6700K / i5 6600K (TDP 91W / Idle 2W)
Tcase (CPU temp) = 64C
Tjunction (Core temp) = 69C
Tj Max (Throttle temp) = 100C
5th Generation 14 nanometer: i7 5775C / i5 5675C (TDP 65W / Idle 2W)
Tcase (CPU temp) = 71C
Tjunction (Core temp) = 76C
Tj Max (Throttle temp) = 96C
4th Generation 22 nanometer: i7 4790K (TDP 88W / Idle 2W)
Tcase (CPU temp) = 74C
Tjunction (Core temp) = 79C
Tj Max (Throttle temp) = 100C
4th Generation 22 nanometer: i5 4690K (TDP 88W / Idle 2W)
4th Generation 22 nanometer: i7 4770K / i5 4670K (TDP 84W / Idle 2W)
Tcase (CPU temp) = 72C
Tjunction (Core temp) = 77C
Tj Max (Throttle temp) = 100C
3rd Generation 22 nanometer: i7 3770K / i5 3570K (TDP 77W / Idle 6W)
Tcase (CPU temp) = 67C
Tjunction (Core temp) = 72C
Tj Max (Throttle temp) = 105C
2nd Generation 32 nanometer: i7 2600K / i5 2500K (TDP 95W / Idle 8W)
Tcase (CPU temp) = 72C
Tjunction (Core temp) = 77C
Tj Max (Throttle temp) = 98C
Previous Generation 45 nanometer: i7 860 / i5 750 (TDP 95W / Idle 12W)
Tcase (CPU temp) = 72C
Tjunction (Core temp) = 77C
Tj Max (Throttle temp) = 100C
Previous Generation 45 nanometer: i7 920 D0 (TDP 130W / Idle 12W)
Tcase (CPU temp) = 67C
Tjunction (Core temp) = 72C
Tj Max (Throttle temp) = 100C
-> Core 2
Legacy 45 nanometer: Q9650 E0 (TDP 95W / Idle 16W)
Tcase (CPU temp) = 71C
Tjunction (Core temp) = 76C
Tj Max (Throttle temp) = 100C
Legacy 65 nanometer: Q6600 G0 (TDP 95W / Idle 24W)
Tcase (CPU temp) = 71C
Tjunction (Core temp) = 76C
Tj Max (Throttle temp) = 100C
Section 8 - Power and Temperature
The previous Sections have explained Intel’s Specifications, and how temperatures are measured and relate to one another. This Section will put the Specifications into perspective, and explain why Thermal Design Power, Tcase and Tj Max sometimes seem to conflict with recommended real-world Core temperatures.
Although Intel measures Tcase on the surface of the Integrated Heat Spreader (IHS), they also calculate the Tcase Specification based on their reference design stock coolers. There are several cooler models with different Thermal Design Power (TDP), which is expressed in Watts.
Certain TDP coolers are packaged with different TDP processors. For example, several Generations of Quad Core processors were packaged with a universal 95 Watt TDP cooler. When this 95 Watt cooler was matched with the 95 Watt i7 2600K, Tcase was calculated and measured at 72C. When the same 95 Watt cooler was packaged with the 77 Watt i7 3770K, Tcase was calculated and measured at only 67C (shown above in Section 7).
The i7 6700K and i5 6600K don’t include a stock cooler. Tcase is instead based on Intel’s new cooler which is sold separately: Intel’s Skylake Cooler - This is what Intel's first CPU cooler for Skylake looks like - Unlike the 95 Watt cooler packaged with earlier Quad Core processors, Intel’s new cooler is 130 Watts for the 91 Watt 6700K and 6600K, so Tcase was calculated and measured at a very low value of 64C (shown above in Section 7).
Tcase Specifications are determined by the stock cooler TDP. This is the primary reason why there’s so much variation in Tcase Specifications among processors of the same Core Count. For example, let’s compare two popular Quad Core CPU’s:
i7 6700K - 4 Cores / 8 Threads / 91 Watts TDP / Cooler 130 Watts TDP / Tcase = 64C.
i7 4790K - 4 Cores / 8 Threads / 88 Watts TDP / Cooler 95 Watts TDP / Tcase = 74C.
As you can see, although there's only a 3 Watt difference between the CPU's, there’s a 35 Watt difference in coolers which results in a 10C difference in Tcase specifications.
Additionally, processors without Hyperthreading (i5’s, Pentiums and Celerons) run cooler than their counterparts with Hyperthreading (i7’s and i3’s), even though they may have the same Tcase and TDP Specifications. The key word in the term “Thermal Design Power” (TDP) is Design. i5’s follow the i7 Design just as Pentiums and Celerons follow the i3 Design. The differences are Core Count, Cache, Instruction Sets and Hyperthreading, the latter of which has a pronounced effect on Core temperatures.
When sustained Core temperatures exceed 80C, some processors may become unstable. Even though Tj Max (Throttle temp) may be 100C for your CPU variant, it’s highly recommended that you avoid running your processor at Core temperatures that begin to approach Throttle temperature due to “Electromigration” (explained below in Section 9).
Core i 6th Generation processors have features such as Configurable TDP (cTDP) and Scenario Design Power (SDP) which may trigger throttling as low as 80C. See Sections 5.1.4 and 5.1.7 under “Thermal Management” in the following document: 6th Generation Intel Processor Datasheet - http://www.intel.com/content/dam/ww...sktop-6th-gen-core-family-datasheet-vol-1.pdf
Intel’s silicon fabrication has been very consistent for the past decade across dozens of variants. Processors with low Tcase specifications are just as thermally capable as the 88 Watt 4th Generation i7 4790K with a Tcase of 74C, which again is CPU temperature, not Core temperature; (Tcase + 5 = Core temperature).
So regardless of your processor’s microarchitecture, TDP, Tcase and Tj Max Specifications, BIOS settings, overclock, hardware configuration, CPU cooler, Ambient temperature, app / game / stress test software workloads or any other variables, here's the bottom line:
Core temperatures in the mid 70’s are safe. Sustained Core temperature greater than 80C is too hot for ultimate stability, performance and longevity.
Section 9 - Overclocking and Voltage
Overclocking is always limited by two factors; voltage and temperature. As Core speed (MHz) is increased above a level unique to each processor (silicon lottery), Core voltage (Vcore) must also be increased to maintain stability. This increases power consumption (Watts) which results in increased Core temperatures.
Overclocked processors using increased Vcore can run up to 50% above TDP. This is why high TDP air or liquid cooling is critical to keep Core temperatures under 80C. Overclocking should not be attempted with Vcore settings in “Auto” because BIOS will apply significantly more voltage than is necessary to maintain stability.
Even when using manual Vcore settings, excessive Vcore and temperatures may result in accelerated "Electromigration" - Google
This prematurely erodes the traces and junctions within the processor's layers and nano-circuits, which will eventually result in blue-screen crashes that become increasingly frequent over time. CPU's become more susceptible to Electromigration with each Die-shrink. However, Intel's advances in FinFET technology has improved the voltage tolerance of their 14 nanometer architecture.
Here’s a list of the maximum recommended Vcore settings:
-> Core i
6th Generation 14 nanometer ... 1.400 Vcore
5th Generation 14 nanometer ... 1.400 Vcore
4th Generation 22 nanometer ... 1.300 Vcore
3rd Generation 22 nanometer ... 1.300 Vcore
2nd Generation 32 nanometer ... 1.350 Vcore
Previous Generation 32 nanometer ... 1.350 Vcore
Previous Generation 45 nanometer ... 1.400 Vcore
-> Core 2
Legacy 45 nanometer ... 1.400 Vcore
Legacy 65 nanometer ... 1.500 Vcore
When tweaking your processor near it's highest overclock, keep in mind that for an increase of 100 MHz, a corresponding increase of about 50 millivolts (0.050) is needed to maintain stability. If 75 to 100 millivolts or more is needed for the next stable 100 MHz increase, it means your processor is overclocked beyond it's capability.
With high TDP air or liquid cooling you might reach the Vcore limit before 80C. With low-end cooling you’ll reach 80C before the Vcore limit. Regardless, whichever limit you reach first is where you should stop and declare victory. Testing is explained in Sections 11 through 14.
Remember to keep overclocking in perspective. For example, the difference between 4.4 GHz and 4.5 Ghz is less than 2.3%, which has no noticeable impact on overall system performance. It simply isn’t worth pushing your processor beyond recommended Core voltage and Core temperature limits just to squeeze out another 100 MHz.
Section 10 - The TIM Problem
Core i 3rd through 6th Generation processors are very sensitive to small increases in voltage and frequency. When overclocked, temperatures might exceed 80C, so high-end air or liquid cooling is critical. 3rd through 6th Generation processors are more difficult to cool than earlier processors for three reasons:
(1) The 3rd and 4th Generation 22 nanometer Die, and the 5th and 6th Generation 14 nanometer Die have significantly less surface area in contact with the underside of the Integrated Heat Spreader (IHS), than the larger 2nd Generation 32 nanometer Die.
(2) 3rd through 6th Generation processors have more transistors packed into a smaller Die than 2nd Generation processors.
(3) 3rd through 6th Generation processors use Thermal Interface Material (TIM) between the top of the Die and the underside of the IHS. Solder, which has superior thermal transfer characteristics, was instead used in 2nd Generation and earlier processors, and is used in Intel's "High End Desktop Processors" - Intel® High End Desktop Processors
http://i1275.photobucket.com/albums/y446/CompuTronix52/Package_zps43993989.jpg(Illustration from Intel Desktop 4th Gen Intel® Core™ Processors Datasheet, Vol. 1, Figure 24).
Since the bonding material which seals the perimeter of the IHS to the Substrate is slightly too thick, this tends to increase the space between the underside of the IHS and the Die, which can cause the TIM to compress unevenly. The effect of this manufacturing procedure is that many processors show a wide deviation between Core temperatures, or one Core which runs much hotter than it's neighbors.
This has encouraged some overclockers to "de-lid" or remove their processor's IHS, which basically involves thoroughly removing the bonding material, replacing only the TIM and then restoring the IHS. Typical results are significantly lower Core temperatures and less deviation between Cores. Here's an excellent YouTube - - that shows before, how-to, and after. Beware that de-lidding will void your warranty, and you can easily damage or destroy your processor.
Intel has addressed these thermal problems in their Haswell refresh. The Devil's Canyon processors have an improved IHS alloy and a new Polymer TIM. Although not as thermally efficient as solder, temperatures have been improved by several degrees.
Regardless, 4th Generation processors differ from their 3rd Generation counterparts in that they have a Fully Integrated Voltage Regulator (FIVR) on the Die, instead of on the motherboard, which increases their Thermal Design Power (TDP). Also, due to a 4.0 GHz clock speed, 4.4 GHz Turbo and increased Vcore, the 4th Generation 88 Watt Devil's Canyon i7 4790K runs hotter at 100% workload than any of it’s predecessors.
Note: 5th Generation 14 nanometer Broadwell processors also have a FIVR on the Die, but the TDP is much lower at only 65 Watts. 6th Generation 14 nanometer Skylake processors do not have Voltage Regulators on the Die. Even though TDP is 91 Watts, thermal behavior is similar to 3rd Generation Ivy Bridge 77 Watt processors.
Section 11 - Thermal Testing Tools
In order to properly test your temperatures, you'll need:
-> A trusted analog, digital or IR thermometer to measure Ambient temperature.
-> The following freeware utilities downloaded and installed -
• Core Temp - http://www.alcpu.com/CoreTemp
• CPU-Z - CPU-Z | Softwares | CPUID
• Prime95 v26.6 - Windows Downloads Center: Prime95 26.6
-> Optional; Install Real Temp (developed for Intel processors) to test your Core temperature sensors or monitor your temperatures - http://www.techpowerup.com/downloads/2089/real-temp-3-70
-> Optional; Install SpeedFan if you’d like to use the “Charts” to see your thermal signatures - http://www.almico.com/sfdownload.php
Section 12 - Thermal Testing Basics
We all remember science class where one of the guiding principles for conducting a controlled experiment, is that it's critical to follow the same procedure every time. This minimizes variables so results will be consistent and repeatable.
Since everyone tests their rigs with different hardware using X stress software at Y Ambient temperatures with Z measuring utilities resulting in CPU or Package or Core temperatures, it's impossible to compare apples to apples. This is why processor temperatures are so confusing.
There are only three relevant values; Ambient temperature, Core temperature at steady-state 100% workload, and Core temperature at dead idle. Applications, rendering, encoding and gaming are partial workloads with fluctuating temperatures which aren’t suitable for thermal testing or accurate temperature comparisons.
Sections 13 and 14 will explain how to properly test your rig at load and idle using standardized methods which minimize hardware, software and environmental variables. Follow the "Setup" in both Sections to replicate Intel's test conditions. Each 10 minute test will establish a valid thermal baseline.
Section 13 - Thermal Testing @ 100% Workload
Prime95 Version 26.6 Small FFT's is the standard for CPU thermal testing, because it's a steady-state 100% workload which runs Core 2 processors and Core i variants with Hyperthreading within 3% TDP at stock settings. This is the test that Real Temp uses to test sensors.
Core i 2nd through 6th Generation CPU's have AVX (Advanced Vector Extension) instruction sets. Recent versions of Prime95 such as 28.9 run AVX code on the Floating Point Unit (FPU) math coprocessor, which produces unrealistically high temperatures. The FPU test in the utility AIDA64 shows similar results.
Prime95 v26.6 produces temperatures on 3rd through 6th Generation processors more consistent with 2nd Generation, which also have AVX instructions, but do not suffer from thermal extremes due to having fewer transistors in a larger Die with greater surface area, and a soldered Integrated Heat Spreader.
Note: Keep in mind that we're thermal testing only. Stability testing is not within the scope of this Guide, which assumes your rig is stable. If you're overclocked, then a combination of stress tests, apps or games must be run to verify CPU stability.
If you’re overclocked and run AVX apps, you may need to reduce Vcore and clock speed and / or upgrade your cooling so that Core temperatures don’t exceed 80C. Asus RealBench runs a realistic AVX workload within 3% TDP at stock settings, however, it’s a cyclic workload for stability testing, which isn’t suitable for CPU thermal testing.
• Asus RealBench - http://rog.asus.com/rog-pro/realbench-v2-leaderboard/
Prime95's default test, Blend, is also a cyclic workload for testing memory stability, and Large FFT's combines CPU and memory tests. As such, Blend and Large FFT's both have cyclic workloads which aren’t suitable for CPU thermal testing.
Other stability tests such as Linpack and Intel Burn Test have cycles that peak at 110% workload, which again aren’t suitable for CPU thermal testing. The test utility OCCT runs elements of Linpack and Prime95, but will terminate the CPU tests at 85C.
The "Charts" in SpeedFan span 13 minutes, and show how each test creates different thermal signatures.
http://i1275.photobucket.com/albums/y446/CompuTronix52/SpeedFanTempGuideGraph_zpsd98effba.jpgShown above from left to right: Small FFT's, Blend, Linpack and Intel Burn Test.
Note the steady-state thermal signatures of Small FFT's, which allows accurate measurements of Core temperatures. A steady-state 100% workload is critical for thermal testing.
http://i1275.photobucket.com/albums/y446/CompuTronix52/SmallFFTsIntelETUAIDA64_zps2b0c9ff0.jpgShown above from left to right: Small FFT's, Intel Extreme Tuning Utility CPU Test, and AIDA64 CPU Test.
Intel Extreme Tuning Utility is also a cyclic workload. Although AIDA64's CPU test is steady-state, the workload is well below TDP, which is insufficient for thermal testing. All other AIDA64 CPU test combinations are cyclic workloads, which again aren't suitable for thermal testing.
Setup:
Testing should be performed with your computer clear of desk enclosures or items that block airflow. Covers should be removed and all fans and circulating pump (if equipped with liquid cooling) at 100% RPM, so temperatures can be tested under ideal conditions.
Testing close to 22C Ambient is preferred so as to provide normal thermal headroom, but is not required. During the summer climate, if adequate A/C isn’t available, then test late at night or early in the morning when Ambient is lowest.
Core temperatures rise and fall with Ambient temperature. When testing above or below 22C, it’s important to "normalize" test results to establish a valid thermal baseline. This minimizes variables so results will be consistent and repeatable.
Summer climates create above normal Ambient which decreases overclocking headroom. For example, if your measured Ambient is 4C above Standard, subtract 4C from your reported Core temperatures to normalize your test results.
Winter climates create below normal Ambient which increases overclocking headroom. For example, if your measured Ambient is 4C below Standard, add 4C to your reported Core temperatures to normalize your test results.
Core temperatures normalized to Standard Ambient are your baseline temperatures. Establishing a baseline is important because as Ambient changes, if you maintain your hardware configuration and BIOS settings, a baseline gives you a consistent point of reference. You can repeat the test whenever you like, to see if your rig is maintaining it’s thermal performance.
Test:
Run Prime95 v26.6 Small FFT's for 10 minutes, then use your thermometer to measure Ambient. Use Core Temp to measure your Core temperatures.
Results:
If reported Core temperatures exceed 80C, you should reduce Vcore and clock speed and / or improve cooling. Core temperatures in the mid 70's are safe.
Intel’s specification for Digital Thermal Sensor (DTS) accuracy is +/- 5C. This means deviations between the highest and lowest Cores may be 10C, so "Average" Core temperature is often more realistic.
On processors with more than 2 Cores, the inner Cores run warmer because they’re insulated by the outer Cores. Here’s the physical layout in 2nd, 3rd and 4th Generation Quad Core processors, and an example of how it typically affects Core temperatures:
IGPU = Not in use (PCIE graphics card in use)
Core #0 = 75C (insulated by IGPU and Core #1)
Core #1 = 78C (insulated by Core #0 and #2)
Core #2 = 76C (insulated by Core #1 and #3)
Core #3 = 71C (insulated by Core #2 only)
Core Average = 75C.
Note: When viewing your temperatures in Core Temp, values which reach 81C or higher will change from black to amber, which indicates caution.
Normalize your results to Standard Ambient and record the values for future reference.
Section 14 - Thermal Testing @ Idle
Look closely at the SpeedFan Charts above, where idle temperatures are shown between load temperatures. Note that some Cores have more "range" than others and idle lower. Sensors can be tested with Real Temp. Core temperature sensors are more accurate at high temperatures for Throttle protection, so idle temperatures may not be accurate.
If "Speedstep", also called Enhanced Intel Speedstep Technology (EIST), is disabled in BIOS, then depending on Vcore and clock speed, idle Power can be nearly 40 Watts, which will result in high idle temperatures, especially when combined with high Ambient temperature.
Setup:
In addition to using the previous Setup in Section 13, Speedstep and all "C" States must be enabled to achieve the lowest possible idle temperatures. Also, if Windows Power Options for "Balanced" or "Power saver" is not set correctly, then Speedstep will not work ... OR ... if Windows Power Options is set to "High performance", then Speedstep will not work because Minimum processor state can‘t be set.
To check this, click on Control Panel, Power Options, then to the right of the selected plan, click on Change plan setting. Next click on Change advanced power settings, then drag the scroll bar down. Click on + next to Processor power management, then click on + next to Minimum processor state. This Setting must be 5%. If it's not, then correct it and click Apply.
Restart into BIOS and confirm that you've saved your settings to a Profile. Next, change all settings to stock (Default / Auto) including SpeedStep, all C States and Vcore, then save and exit. Reboot into Windows and confirm that your rig is at dead idle; no programs running, and off line. No Folding or SETI or "tray-trash" running in the background, and less than 3% CPU Usage under the "Performance" tab in Windows Task Manager.
Use CPU-Z to confirm that Core Voltage and Core Speed has decreased as follows:
-> Core i
6th Generation 14 nanometer ... about 0.8 Volts @ 800 MHz
5th Generation 14 nanometer ... about 0.8 Volts @ 800 MHz
4th Generation 22 nanometer ... about 0.8 Volts @ 800 MHz
3rd Generation 22 nanometer ... about 0.9 Volts @ 1600 MHz
2nd Generation 32 nanometer ... about 1.0 Volts @ 1600 MHz
Previous Generation 32 nanometer ... about 1.0 Volts @ 1600 MHz
Previous Generation 45 nanometer ... about 1.0 Volts @ 1600 MHz
-> Core 2
Legacy 45 nanometer ... about 1.1 Volts @ 2000 MHz
Legacy 65 nanometer ... about 1.25 Volts @ 1600 MHz
Use Core Temp to confirm that Power has decreased as follows:
-> Core i
6th Generation 14 nanometer ... about 2 Watts
5th Generation 14 nanometer ... about 2 Watts
4th Generation 22 nanometer ... about 2 Watts
3rd Generation 22 nanometer ... about 6 Watts
2nd Generation 32 nanometer ... about 8 Watts
Previous Generation 32 nanometer ... about 8 Watts
Previous Generation 45 nanometer ... about 12 Watts
Note: Idle Volts and Watts may differ depending on BIOS versions and motherboard models. Power (Watts) isn't measured on Previous Generation Core i Socket 1366 variants and Legacy Core 2 processors, but for general reference, idle power for several popular CPU's is shown in Section 7 - Relative Temperatures.
Test:
Allow your rig to "settle" for 10 minutes, then use your thermometer to measure Ambient. Use Core Temp to measure your Core temperatures.
Results:
Core i 2nd through 6th Generation processors should idle at less than 8C above Ambient. This means at 22C Standard Ambient your Cores should idle just under 30C. Certain Previous Generation Core i variants and Legacy Core 2 processors may idle several degrees higher. Better cooling and lower idle power produce lower idle temperatures.
Normalize your results to Standard Ambient and record the values for future reference. When finished testing, restore your system to it's previous configuration.
Section 15 - Improving Temperatures
Whether your computer is a stock workstation or an overclocked gaming rig, achieving the lowest possible temperatures always depends on components, configuration and airflow. Here's a few thoughts:
• Intel coolers are barely adequate at stock. If you want to overclock then upgrade your cooler.
• Use Manual Vcore settings. Auto applies excess voltage which means more power and heat.
• Memory overclock and XMP Profiles can cause Core i processors to run a few degrees hotter.
• Axial flow graphics cards recirculate heat. Linear flow cards exhaust heat from the case.
Examples:
Axial - http://www.newegg.com/Product/Product.aspx?Item=N82E16814487248
Linear - http://www.newegg.com/Product/Product.aspx?Item=N82E16814487247
• Axial cards work well with a liquid cooled CPU. Linear cards work well with an air cooled CPU.
• SLI / CrossFire works best with Linear cards. Axial cards dump massive heat in your case.
• A hot case stresses hard drives, memory, chipsets, voltage regulators and power supply.
• High performance computers need unrestricted airflow in and out, so location is critical.
• Load temperatures that drop over a degree or two with case covers off means poor airflow.
• Good cable management creates good airflow. Use zip-ties, patience and attention to detail.
• Quality fans are important, but if you want a quiet computer then consider a fan controller.
• If your case just doesn't breathe well, then perhaps it's time to upgrade to one that does.
• If your rig runs 24/7, then hard drive and fan bearings are wearing, and dust is accumulating.
• Clean the dust out of your rig. Perform regular Planned Maintenance Inspections (PM's).
• Replace your TIM. Most Thermal Interface Material typically begins to fail after 2 years.
Thermal Interface Material (TIM):
Thermal Paste Comparison, Part One: Applying Grease And More - http://www.tomshardware.com/reviews/thermal-paste-heat-sink-heat-spreader,3600.html
Thermal Paste Comparison, Part Two: 39 Products Get Tested - http://www.tomshardware.com/reviews/thermal-paste-performance-benchmark,3616.html
The proper installation of Intel's stock cooler:
Intel Stock Cooler Installation Guide - http://www.tomshardware.com/forum/338655-28-intel-stock-cooler-installation-guide
Choosing an aftermarket cooler:
Air Cooling vs Water Cooling : Things You Need To Know - http://www.tomshardware.com/forum/id-2196038/air-cooling-water-cooling-things.html
Alternatives to the Hyper 212+/Evo for budget cooling - http://www.tomshardware.com/forum/id-2705157/alternatives-hyper-212-evo-budget-cooling.html
Section 16 - Summary
• Standard Ambient temperature is 22C.
• Ambient affects all computer temperatures.
• No temperatures can be less than or equal to Ambient.
• As Ambient increases, thermal headroom decreases.
• BIOS or CPU temperature may not be accurate.
• Package temperature is the hottest Core.
• Core temperature is the standard for thermal measurement.
• Core temperatures respond instantly to changes in load.
• 80C sustained Core temperature is too hot.
• Core temperatures in the mid 70's are safe.
• Excessive Vcore and temperatures accelerate electromigration.
• Prime95 v26.6 Small FFT's is the standard for thermal testing.
• Deviations between highest and lowest Cores may be 10C.
• Core temperature sensors are more accurate at high temperatures.
• Idle temperatures may not be accurate.
• Sensors can be tested with Real Temp.
