• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Heatsink Temperature of 2GB Memory Modules

Asher Kelman

OPF Owner/Editor-in-Chief
Rob-Art at Barefeats.com has been studying the temperature of the vicinity of RAM for the new Intel Macs. Well not actually the chip itself, but the heatsink. He looks at the genuine Apple modules and replacements with different heat sinks by other companies.

http://www.barefeats.com/quick.html

Excerpted: "HEAT SINK TEMPERATURE TESTS........

We can say that ALL brands with "advanced finned" heat sinks ran cooler than Apple's factory memory with its advanced heat sink."

My emphasis.

This is worth looking into. Rob also discusses the placement of the chips in the upper and lower tiers and the how it affects temperatire. So with less RAM, perhaps they should perhaps be placed on the lower tier where temperature was not above 150F compared to the upper tier where on the left side, it was 30F higher!

These tests are still in progress and it is is worth keeping an eye on final results!

Asher
 

Don Lashier

New member
Quote from the website: "We are not equipped to sample temperatures on the chips. We only can sample temperatures at various spots on the heat sink."

Thus is it not clear to me which is better. Conceivably a hotter heatsink could mean that it is conducting heat from the chip better than a heatsink which measures cooler. Or it could mean that the hotter heatsink is not radiating the heat as well to the air. The only way to know for sure is to measure the chip temperature.

- DL
 

Asher Kelman

OPF Owner/Editor-in-Chief
Don, there is some logic in your statement, but I still think that we can take the heat sink temperatures do in fact reflec variances in chip temperatures.

Heat travels in a gradient until there is no gradient. The mass of the heat sink just takes away more heat. The fins provide a greater surface area for cooling air to remove that heat.

Given a sufficient attachment for the full area of the chip, the heat sink will not be hotter because it conducts heat away better. Of course, if by error, the heat sink was designed with some intermediate insulation, then of course, a lower temperature could follow. However, I can't imagine that this could be so.

Further, a fin system with more surface area would intuitively fit in with the observed lower temperature, so I have no problem believing the improvement is due to the design of these new heat sinks. Of course, we must assume that the chips themselves are exchangable.

Anyway, my working hypothesis is that the heat sink has a lower temperature than the chip and its temperature reflects the temperature of the chip or else it would not work and if it did, that would be a new chapter in physics.

Asher
 

Don Lashier

New member
Asher, most likely the difference is due to a better fin design but note that the thermal conductivity/resistance of the chip/sink interface varies by a factor of about 1:10 just based on the goo/glue used - all other things being equal. The uniformity (flatness) of the heatsink interface is also extremely important which is why overclockers go to all the trouble to lap their heatsinks. Also note this quote from overclockers.com
A TIM (Thermal Interface Material) joint is an element of every heatsink or waterblock installation, and is also the most variable part of any CPU cooling system. Not always appreciated is that the TIM joint variation can easily exceed the difference between different waterblocks, so its control is essential for comparative testing.

Intel chips have internal thermometers which can be read out. I assume that the fbdimms do also since apparantly the system will detect overheating and slow memory operations to prevent damage. There must be a way to read out these registers and this would be a much better indicator of efficiency. Note that the Apple mem also has fins and it's hard to determine from the photos what the differences are.

- DL
 

Asher Kelman

OPF Owner/Editor-in-Chief
It could be that there is already software to read the actual temperature of the module itself.

Your point on the cement is well taken. After all it is not as if we know they are all using the same cement/thickness etc.

Maybe the test platform for this is on a PC!

Asher
 

Stan Jirman

New member
Note that the order in which you install the memory matters. You can't simply start adding RAM to the bottom just because you prefer it (because it's cooler there) - that won't work. You have to go 2-top, 2 top / 2 bottom, 4 top / 2 bottom, 4 + 4. You have no choice over this; it simply won't work otherwise.

I have 2x1GB Kingston in my Mac Pro, in addition to 4x512 Apple memory, and no probs with it. Two more 1G Kingstons coming in 2 weeks when I come back from Africa :)
 
Top