Intel Patents Related Component Cooling Concept

Intel was granted a patent described as "extended thermal management" that is based on the idea that the reduction of the temperature of one component can also reduce the temperature of other devices.

The document specifically refers to hardware that is used in both desktop and mobile computer systems, but runs into heat dissipation limitations on notebooks. Depending on user preference, the patent outlines a "thermal zone" consisting of multiple components whose temperature can be controlled and as well as devices that cannot be controlled.

Since the temperature of individual devices can be significantly affected by the temperature of nearby devices, Intel says that "the cooling of a hot controllable component may indirectly help cool a nearby non-controllable component." A thermal zone would be defined by the characteristics of the individual devices that would allow the system to create a "thermal relationship table", which provides the necessary data to achieve indirect cooling of components that cannot be directly controlled to reduce their heat dissipation.

According to Intel, the table would include information "how much change [for] one component [is necessary] in order for that change to have an effect on another component" and "how long it takes for a change in one component to affect another component."

You need to be logged to post a comment

There are 18 Comments. B

Top Comments

14

rex86 , March 15, 2012 9:50 PM

please stop patenting basic ideas. it's not worth it and you're making it worse for everyone.
12

husker , March 15, 2012 9:24 PM

An interesting idea, but I'm not sure if this doesn't just fall into the "common sense" category. All it is doing is using general thermal readings to adjust the output of a specific device.
11

COLGeek , March 15, 2012 10:06 PM

Patents = litigation = money! The new business model for the tech industry.....

Other Comments

12

husker , March 15, 2012 9:24 PM

An interesting idea, but I'm not sure if this doesn't just fall into the "common sense" category. All it is doing is using general thermal readings to adjust the output of a specific device.
-2

theuniquegamer , March 15, 2012 9:30 PM

Good concept but is a little impractical for a enthusiast level desktops because they can afford cooling for almost every components without need to worry about performance
14

rex86 , March 15, 2012 9:50 PM

please stop patenting basic ideas. it's not worth it and you're making it worse for everyone.
11

COLGeek , March 15, 2012 10:06 PM

Patents = litigation = money! The new business model for the tech industry.....
3

jaber2 , March 15, 2012 10:28 PM

Now no one can ever use the terms "extended thermal management" without paying Intel.
4

Tab54o , March 15, 2012 10:41 PM

This is ridiculous.
5

dudzcom , March 15, 2012 11:09 PM

sounds like the just patented Thermal Conductivity
2

HSuke , March 15, 2012 11:27 PM

dudzcomsounds like the just patented Thermal Conductivity

I think you meant that they just patented Heat Convection.
2

Lord Captivus , March 15, 2012 11:29 PM

dudzcomsounds like the just patented Thermal Conductivity

I was about to write the SAME thing!
0

ava__ , March 15, 2012 11:33 PM

theuniquegamerGood concept but is a little impractical for a enthusiast level desktops because they can afford cooling for almost every components without need to worry about performance

probably they can afford earplugs too
2

TeraMedia , March 15, 2012 11:43 PM

Actually, this makes a lot of sense. Let's say you increase the voltage on your DRAM so that you can OC it. That means your DRAM is hotter than it would otherwise typically be. Then lets say that you run an algorithm that is hard on your DRAM , but doesn't tax the CPU all that much. You end up with an isolated hot-spot in your DRAM. Your CPU fan speed would still be slow due to the low load on the CPU. Your case fans might not have sped up very much because the total system temperature might not have gone up by much at all. But the additional heat from your DRAM could be enough to cause an instability.

Now lets say that the CPU fan controller is configured so that it knows there is a relationship between CPU fan speed and DRAM temperature. When this scenario happens, the CPU fan controller senses the increased local temperature of the DRAM, and spins up the CPU fan to help maintain system stability.

Similar concepts could probably apply to Vregs, chipsets, PCIe switches, Ethernet PHY chips, and other system components that typically don't have temperature management logic built-in. It wouldn't be the most noise-efficient way to keep such components adequately cooled, but it would probably be the most economical and would certainly be better than spinning available fans at max speed all the time.
-1

spookyman , March 16, 2012 12:13 AM

I thought called this a thermistor? If said item got to hot, it shuts off the processor.
0

Dogsnake , March 16, 2012 3:02 AM

I understand what they are saying. How is it patentable? Now if they started to manufacture their MB with defined thermal zones with interlocking sensors and cooling solutions it might lead to some future design for total MB cooling.
-6

anonymous@guest , March 16, 2012 3:33 AM

Have you any idea how small is 1 CPU? Less that the half head of a needle and the big question for Intel is how managed to make produce so much heat with so little volts, I mean our water heater is not that efficient!?

Nobody cares about their crap.
1

memadmax , March 16, 2012 4:38 AM

meh...

Didn't know that a flowchart was patentable...

=P
1

A Bad Day , March 16, 2012 5:14 AM

dudzcomsounds like the just patented Thermal Conductivity

Actually, I can one-up their patent by patenting thermal physics, and suing anyone who dares to take advantage of the physics.

Screw prior art!
-1

rooni , March 16, 2012 10:39 AM

TeraMediaActually, this makes a lot of sense. Let's say you increase the voltage on your DRAM so that you can OC it. That means your DRAM is hotter than it would otherwise typically be. Then lets say that you run an algorithm that is hard on your DRAM , but doesn't tax the CPU all that much. You end up with an isolated hot-spot in your DRAM. Your CPU fan speed would still be slow due to the low load on the CPU. Your case fans might not have sped up very much because the total system temperature might not have gone up by much at all. But the additional heat from your DRAM could be enough to cause an instability.Now lets say that the CPU fan controller is configured so that it knows there is a relationship between CPU fan speed and DRAM temperature. When this scenario happens, the CPU fan controller senses the increased local temperature of the DRAM, and spins up the CPU fan to help maintain system stability.Similar concepts could probably apply to Vregs, chipsets, PCIe switches, Ethernet PHY chips, and other system components that typically don't have temperature management logic built-in. It wouldn't be the most noise-efficient way to keep such components adequately cooled, but it would probably be the most economical and would certainly be better than spinning available fans at max speed all the time.

False. Also with a half-decent case case you can run fans at max and not hear them over your GPU while having your system stable at 100% load 24/7
1

techcurious , March 16, 2012 4:19 PM

I can't believe this is actually patentable! Sometimes when I'm going to be gaming for a while, I decide to lower the room temperature by a few degrees (room equiped with A/C) so my PC won't heat up as much while I'm gaming. I should go patent this concept immediately even though I am sure most gamers have been aware of this solution long before I ever played my first video game!