Peter, you need to review basic thermodynamic laws and restrictions.
I can do quantum mechanics, I think I can manage thermodynamics.
Keith Myers wrote:
Area of thermal transfer interface is integral to calculations of heat transfer. As Ian points out the speed of transfer is directly related. It is easier to remove more heat per unit of time with a larger surface area. To get an idea of the relative size of the IHS of Epyc/Threadripper compared to a Ryzen IHS view this short YT video.
It is related yes. But the whole heatsink setup is a series of temperature drops, just like resistors in series. The "resistor" that is the contact between the CPU and heatsink is a very low resistance compared to the fin-to-air "resistor". It's like a circuit with 1 ohm, 30 ohms, 30 ohms in series. Changing the 1 to 0.25 ohms doesn't do much.
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.
My 2nd attempt will include removing a Cd-Rom drive
I don't mean to be rude but that had me in fits of laughter. Do they still make those things? I don't have CD/DVD/Blu-ray in any of my machines. What's it for?
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.
I once had a teacher who said: "If you don't have the time to do something right the first time, then don't do it at all".
Hence, I have taken that statement to heart and literally always made the correct choices in whatever I have done. It may take me a while longer, but I do get the job done... correctly.
That saying is just silly. What if you have to drive to your destination (some very important family matter), and only have time to do a shoddy fix on your engine? Do you do that and risk damage, or do you not go?
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.
My 2nd attempt will include removing a Cd-Rom drive
I don't mean to be rude but that had me in fits of laughter. Do they still make those things? I don't have CD/DVD/Blu-ray in any of my machines. What's it for?
For me it's because I get dvd disks cheap, people throw them away all the time now and it's cheaper than buying usb sticks for OS installs. MOST of my pc's do not have one but I do have a couple of portable DVD drives that plug into a couple of usb ports. I have a stack of internal dvd drives in a couple of drawers but they will hit the trash at some point, I've already trashed the non sata ones.
We are receiving our Threadripper 5995wx Dell workstation next week. The CPU is 280 watts, air cooled. I am very interested to see the thermals/noise when pushed to 100%, 24/7.
For me it's because I get dvd disks cheap, people throw them away all the time now and it's cheaper than buying usb sticks for OS installs.
There's a certain person selling USB sticks on Ebay :-) Only 10 dollars for 32GB.
mikey wrote:
MOST of my pc's do not have one but I do have a couple of portable DVD drives that plug into a couple of usb ports. I have a stack of internal dvd drives in a couple of drawers but they will hit the trash at some point, I've already trashed the non sata ones.
I never trash anything. Sell it, give it away on freecycle, take it apart and use the laser....
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.
We are receiving our Threadripper 5995wx Dell workstation next week. The CPU is 280 watts, air cooled. I am very interested to see the thermals/noise when pushed to 100%, 24/7.
I will be happy to post that info when I have it.
I have worked in a school and we also bought Dell computers exclusively. Decent price and well made, although non-standard parts sometimes. I ran Boinc on a few hundred of them and I did receive a few complaints about noise. Apparently 20 computers in one classroom running flat out is loud. But most of the teachers didn't mind, the pupils were louder. I fitted a huge TV in the lunch hall and had to add a couple of kW RMS of speakers so they could hear MTV above the racket they made.
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.
Take one off and see how much gunk was sandwiched between the CPU and heatsink, it's a fraction of a mm. And I use SYY 15.7W/mK paste. At a quarter of a mm, that works out to a temperature drop of 0.943C! Even 1mm at 240W with the surface area of my chip would be would be 7.544C. Insignificant. The important things are the fin surface area, the effectiveness of the heat pipes, and the speed of airflow.
you’re missing the point. The comment about TIM was only a rebuttal to your claim that it was metal to metal contact. You’ve proceeded to run with that to make a nice little strawman here.
the point was that the heat is distributed over a larger area. There is not just 1 or 2 dies being cooled pumping out 240W, there are 9 chips on the package sharing that load. Eight CCDs and the IO die. I don’t remeber off hand how much power the IO die pulls, maybe like 20W or so. So that’s 220W across the eight CCDs, and less than 30W each. They are grouped on the substrate in pairs at each quadrant, so you can think of it like cooling 4 60w chips.
you’re confusing heat with temperature. The heat being transferred is moderately high, but low temperatures aren’t hard to achieve because that heat is distributed over a larger area (the IHS) and the cooler is still capable of removing that amount of heat from the IHS.
the heatsink 5 heat pipes, has 56x fins, each fin with approx. 6000 sq.mm area, spaced at a pretty tight 1.5mm. The 92mm fan spins at 3800rpm.
Take one off and see how much gunk was sandwiched between the CPU and heatsink, it's a fraction of a mm. And I use SYY 15.7W/mK paste. At a quarter of a mm, that works out to a temperature drop of 0.943C! Even 1mm at 240W with the surface area of my chip would be would be 7.544C. Insignificant. The important things are the fin surface area, the effectiveness of the heat pipes, and the speed of airflow.
you’re missing the point. The comment about TIM was only a rebuttal to your claim that it was metal to metal contact. You’ve proceeded to run with that to make a nice little strawman here.
It's pretty much metal to metal, since I showed you the calculations for the gunk which is a very thin layer which just fills in any gaps when they aren't perfectly aligned.
Ian&Steve C. wrote:
the point was that the heat is distributed over a larger area.
Which is irrelevant when that is not the highest resistance of heat transfer. Metal to air is the highest.
Ian&Steve C. wrote:
you’re confusing heat with temperature.
I'm doing nothing of the sort. And I object to you thinking I'm confused. This seems to be an American insult, to say "you're confusing x with y". I find it hard to believe you actually think I don't understand the difference between two entirely different things.
Ian&Steve C. wrote:
The heat being transferred is moderately high, but low temperatures aren’t hard to achieve because that heat is distributed over a larger area (the IHS) and the cooler is still capable of removing that amount of heat from the IHS.
Again, the problem is metal to air, solved only by the fins and the fan.
Ian&Steve C. wrote:
the heatsink 5 heat pipes, has 56x fins, each fin with approx. 6000 sq.mm area, spaced at a pretty tight 1.5mm. The 92mm fan spins at 3800rpm.
Then you'd better explain to Zalman the heatsink I've got is a piece of excrement. Either there aren't enough fins or the fans aren't fast enough. Why not make my one as well as yours? Then my 24 thread CPU could run at less than 85C.
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 wrote: Peter,
)
I can do quantum mechanics, I think I can manage thermodynamics.
It is related yes. But the whole heatsink setup is a series of temperature drops, just like resistors in series. The "resistor" that is the contact between the CPU and heatsink is a very low resistance compared to the fin-to-air "resistor". It's like a circuit with 1 ohm, 30 ohms, 30 ohms in series. Changing the 1 to 0.25 ohms doesn't do much.
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.
Tom M wrote:My 2nd attempt
)
I don't mean to be rude but that had me in fits of laughter. Do they still make those things? I don't have CD/DVD/Blu-ray in any of my machines. What's it for?
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.
GWGeorge007 wrote:I once had
)
That saying is just silly. What if you have to drive to your destination (some very important family matter), and only have time to do a shoddy fix on your engine? Do you do that and risk damage, or do you not go?
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.
Peter Hucker wrote: Tom M
)
For me it's because I get dvd disks cheap, people throw them away all the time now and it's cheaper than buying usb sticks for OS installs. MOST of my pc's do not have one but I do have a couple of portable DVD drives that plug into a couple of usb ports. I have a stack of internal dvd drives in a couple of drawers but they will hit the trash at some point, I've already trashed the non sata ones.
We are receiving our
)
We are receiving our Threadripper 5995wx Dell workstation next week. The CPU is 280 watts, air cooled. I am very interested to see the thermals/noise when pushed to 100%, 24/7.
I will be happy to post that info when I have it.
mikey wrote:For me it's
)
There's a certain person selling USB sticks on Ebay :-) Only 10 dollars for 32GB.
I never trash anything. Sell it, give it away on freecycle, take it apart and use the laser....
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.
Boca Raton Community HS
)
I have worked in a school and we also bought Dell computers exclusively. Decent price and well made, although non-standard parts sometimes. I ran Boinc on a few hundred of them and I did receive a few complaints about noise. Apparently 20 computers in one classroom running flat out is loud. But most of the teachers didn't mind, the pupils were louder. I fitted a huge TV in the lunch hall and had to add a couple of kW RMS of speakers so they could hear MTV above the racket they made.
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.
Peter Hucker
)
you’re missing the point. The comment about TIM was only a rebuttal to your claim that it was metal to metal contact. You’ve proceeded to run with that to make a nice little strawman here.
the point was that the heat is distributed over a larger area. There is not just 1 or 2 dies being cooled pumping out 240W, there are 9 chips on the package sharing that load. Eight CCDs and the IO die. I don’t remeber off hand how much power the IO die pulls, maybe like 20W or so. So that’s 220W across the eight CCDs, and less than 30W each. They are grouped on the substrate in pairs at each quadrant, so you can think of it like cooling 4 60w chips.
you’re confusing heat with temperature. The heat being transferred is moderately high, but low temperatures aren’t hard to achieve because that heat is distributed over a larger area (the IHS) and the cooler is still capable of removing that amount of heat from the IHS.
the heatsink 5 heat pipes, has 56x fins, each fin with approx. 6000 sq.mm area, spaced at a pretty tight 1.5mm. The 92mm fan spins at 3800rpm.
_________________________________________________________________________
Ian&Steve C. wrote:Peter
)
It's pretty much metal to metal, since I showed you the calculations for the gunk which is a very thin layer which just fills in any gaps when they aren't perfectly aligned.
Which is irrelevant when that is not the highest resistance of heat transfer. Metal to air is the highest.
I'm doing nothing of the sort. And I object to you thinking I'm confused. This seems to be an American insult, to say "you're confusing x with y". I find it hard to believe you actually think I don't understand the difference between two entirely different things.
Again, the problem is metal to air, solved only by the fins and the fan.
Then you'd better explain to Zalman the heatsink I've got is a piece of excrement. Either there aren't enough fins or the fans aren't fast enough. Why not make my one as well as yours? Then my 24 thread CPU could run at less than 85C.
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.
Sounds like you’ve got it all
)
Sounds like you’ve got it all figured out then.
_________________________________________________________________________