Motherboard and System Reviews

Tom M
Tom M
Joined: 2 Feb 06
Posts: 6381
Credit: 9460616791
RAC: 17297256

Peter Hucker wrote: Keith

Peter Hucker wrote:

Keith Myers wrote:
You can't use them to measure body temp.  The IR reflectance of skin does not match what the sensor is set up for.

It claims anything that isn't shiny is ok.  And they're what's used to do a quick check for covid - on the forehead.  Although I was obviously using it under my tongue.  Mind you a normal mercury thermometer under my tongue shows 31 too.  It correctly reads my arm temperature.

Hey Peter,

Could you offer a review of how good the "under your tongue" motherboard is?

 

 

 

 

ROFLing...

Tom M

A Proud member of the O.F.A.  (Old Farts Association).  Be well, do good work, and keep in touch.® (Garrison Keillor)  I want some more patience. RIGHT NOW!

Tom M
Tom M
Joined: 2 Feb 06
Posts: 6381
Credit: 9460616791
RAC: 17297256

Here is a question based on

Here is a question based on CPU crunching.   I have a b459-F and an X570 motherboard.  If I use the exact same memory modules and CPU on both would the X570 crunch significantly faster?  Context is Universe at Home.

Tom M

A Proud member of the O.F.A.  (Old Farts Association).  Be well, do good work, and keep in touch.® (Garrison Keillor)  I want some more patience. RIGHT NOW!

Ian&Steve C.
Ian&Steve C.
Joined: 19 Jan 20
Posts: 3928
Credit: 45857922642
RAC: 64058330

Keith Myers wrote: Ian&Steve

Keith Myers wrote:

Ian&Steve C. wrote:

unfortunately the Asrock boards do not seem to have CPU VRM thermal sensors or dont have the readings available in the IPMI. so they are unknown to me. 

The VRM temps on the Asrock Rack boards are exposed by sensors and the standard nct6775 kernel module as Sensor 2.  Easy enough to verify with a IR heat gun pointed at the VRM heat sink.

I do not believe this is actually the VRM temp. 35C is abnormally low temp for an anemic CPU VRM running 200W. i think this is a temperature at some location on the motherboard, not near anything generating a lot of heat.

_________________________________________________________________________

Tom M
Tom M
Joined: 2 Feb 06
Posts: 6381
Credit: 9460616791
RAC: 17297256

Ian&Steve C. wrote: but 240+

Ian&Steve C. wrote:

but 240+ TDP chips could be a problem I think.

Power table including 7002 cpus  So it looks like a 7702P will be out of my heat range.

But this table seems to imply that a lot of the gen 2 high core count cpus (including the 7702p) could be used (16c-64c)  And certainly the 7601.

Tom M

A Proud member of the O.F.A.  (Old Farts Association).  Be well, do good work, and keep in touch.® (Garrison Keillor)  I want some more patience. RIGHT NOW!

Mr P Hucker
Mr P Hucker
Joined: 12 Aug 06
Posts: 838
Credit: 519032078
RAC: 11160

Tom M wrote:Hey Peter, Could

Tom M wrote:

Hey Peter,

Could you offer a review of how good the "under your tongue" motherboard is?

ROFLing...

Tom M

Oh no, it's the off topic brigade....

If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.

Mr P Hucker
Mr P Hucker
Joined: 12 Aug 06
Posts: 838
Credit: 519032078
RAC: 11160

Ian&Steve C. wrote:I do not

Ian&Steve C. wrote:

I do not believe this is actually the VRM temp. 35C is abnormally low temp for an anemic CPU VRM running 200W. i think this is a temperature at some location on the motherboard, not near anything generating a lot of heat.

My IR sensor specifically says it's inaccurate on anything shiny, which I assume the VRM is.

If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.

Keith Myers
Keith Myers
Joined: 11 Feb 11
Posts: 4957
Credit: 18625424682
RAC: 5464114

Ian&Steve C. wrote: I do not

Ian&Steve C. wrote:

I do not believe this is actually the VRM temp. 35C is abnormally low temp for an anemic CPU VRM running 200W. i think this is a temperature at some location on the motherboard, not near anything generating a lot of heat.

Nope.  This is the actual temp of my VRM heatsink on both my Epyc mobos as verified with my IR temp gun.  I can cool or heat the heatsink up with cool spray or a heat gun and the Sensor 2 temp tracks the actual temp verified with the IR temp gun.

My IR temp gun has a laser pointer and it is easy to target the location and know that you are reading the actual temp of the heatsink. You can put the temp gun right on top of the heatsink and know that is what you are really reading and not some background value.

Or I can verify just as easily with a Type K thermocouple in physical contact with the heatsink and hooked up to my VRM.

 

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6585
Credit: 310935802
RAC: 78054

Peter Hucker wrote:Although

Peter Hucker wrote:

Although alarmingly it agrees with every other thermometer that my body temperature is 31C.

It means that you are a cool guy ! ;-)

I use a hand held infrared radiometer with my work. Ideally it can read the temperature of one's eardrum when correctly directed into an ear canal, where it usually becomes an excellent proxy for body core temperature and a far more polite method than the truly accurate one. This device is calibrated very carefully to be accurate for temps around normal body temperature. It costs hundreds of dollars.

But I have found uses in comparing say differences in the heat of two patches of skin : while the readout may not be accurate at, say, 25o C it will still correctly order the temperatures of a sequence of hot objects ( but it does not work when presented with hot coffee alas ).

Interestingly it is different conceptually than traditional thermometric gadgets, which are assumed to reach equilibrium with the object it contacts. These IR gadgets report a temperature based on radiance received but assume that the object pointed at is in thermal equilibrium. This is a subtle difference not normally mattering much, as it depends on what you're up to. For any extra careful use one needs to record the spectrum of infrared radiation - to confirm that the object is sufficiently close to thermal equilibrium as per a black body model - and then project a result from known calibration data.

There are certain areas of (popular) science where all these concerns are muzzed over with hand waving arguments, and I cringe when I see it. Suffice to say : how you actually go about measuring the temperature of something does matter, as not all methods are equal. Most things in our immediate Universe are not at thermal equilibrium, which is good because interesting things ( eg. human life ) happen there.

Max Planck is a man to remember. He worked out what the spectral form of radiation should be from a body in thermal equilibrium and broke a theoretical log jam of late 19th century physics. This ushered in the 'Quantum Age' which is what I think historians will label it down the track. Did you know that he was originally employed to work out the efficiency of newly invented light bulbs, a classic example of un-intended consequences of research. He is also less famously known to have said 'no' to a certain dictator and yet continue on alive ..... to die of natural causes at 89 years old.

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Keith Myers
Keith Myers
Joined: 11 Feb 11
Posts: 4957
Credit: 18625424682
RAC: 5464114

I think we can pretty much

I think we can pretty much agree that our VRM heatsinks are at thermal equilibrium in a closed case under 95-100% utilization, 24-7 usage on BOINC cpu crunching projects like Universe or TN-Grid.

As long as you waited a sufficient amount of time, 15-30 minutes typically, you can assume you have reached thermal equilibrium of the VRM heatsink.

But if using water cooling of a cpu and a large cooling loop capacity, it takes 45-60 minutes before my loop and cpu temps reach equilibrium.

 

Ian&Steve C.
Ian&Steve C.
Joined: 19 Jan 20
Posts: 3928
Credit: 45857922642
RAC: 64058330

Keith Myers wrote: Ian&Steve

Keith Myers wrote:

Ian&Steve C. wrote:

I do not believe this is actually the VRM temp. 35C is abnormally low temp for an anemic CPU VRM running 200W. i think this is a temperature at some location on the motherboard, not near anything generating a lot of heat.

Nope.  This is the actual temp of my VRM heatsink on both my Epyc mobos as verified with my IR temp gun.  I can cool or heat the heatsink up with cool spray or a heat gun and the Sensor 2 temp tracks the actual temp verified with the IR temp gun.

My IR temp gun has a laser pointer and it is easy to target the location and know that you are reading the actual temp of the heatsink. You can put the temp gun right on top of the heatsink and know that is what you are really reading and not some background value.

Or I can verify just as easily with a Type K thermocouple in physical contact with the heatsink and hooked up to my VRM.

first, as heatsinks are heat dissipators, the temperature of the heatsink is lower than the temperature of the component itself. and the laser on IR temp guns is just a guide, it's not reading the temp of that pinpoint laser. it actually has a fairly large swath that it takes a reading from. the reading area at 8" distance is gonna be about a 1-2" circle. so a small component like a VRM, even when it's not obfuscated by a heatsink, will get averaged with its cooler surroundings.

https://www.firecraftsafety.com/pdf/IR%20Thermometers%20explained.pdf

the system reports 39C, yet the VRM heatsink was burning hot to the touch and only tolerable to hold onto for a second. so it's definitely not the right reading. 39C component temp would equal lower heatsink temp and even 39C is only warm to touch with your fingers.

your test about cool and warm spray would yield the same results for the sensor nearby on the board since hot and cool air will migrate around the board. your sprays affected the temp reading, maybe, but not because of the affect on the VRM heatsink IMO.

35C is just not a realistic value for the amount of power and level of VRM we're talking about here.

_________________________________________________________________________

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.