6 ms·

Maybe RAM will finally get more than a 4mm thermal pad and a random bit of Alu for cooling. Seems like most cooling designs have treated RAM as even more of an

Maybe RAM will finally get more than a 4mm thermal pad and a random bit of Alu for cooling. Seems like most cooling designs have treated RAM as even more of an after thought than VRMs up until recently. Even in most servers the accommodation for RAM cooling has basically just been orientating the DIMMs to line up with airflow. They are still packed together with minimal clearance.

AshamedCaptain 4y ago
I am not sure how much this is related to external cooling versus actually internal thermal dissipation. DDR JEDEC standards have actually decreased power consumption on every generation.
- __alexs 4y ago
  They have reduced voltage but power consumption per sq-mm has gone up with increased densities. Many people run DRAM at above JEDEC speeds which usually requires higher voltages too. Peak power consumption of DDR4 is around 375mW/GB @ 1.2V, DDR5 drops this about 10% but also increases the maximum density of a DIMM by 8x to 512GB which is like, 150W for a single DIMM.
  - formerly_proven 4y ago
    There are only three (tiny) 12 V power pins on a DDR5 module, neither that nor the form factor allows for dissipating anywhere close to 150 W. The teased 512 GB Samsung module doesn't even have a heatspreader.
    - __alexs 4y ago
      VIN_BULK is 5V with a max current of 2A but every data pin provides current that is used on the DIMM in some respect.
      - formerly_proven 4y ago
        I thought they moved to 12 V, thanks for the correction!
        __alexs 4y ago
        I'm just going by the Micron UDIMM data sheets. https://media-www.micron.com/-/media/client/global/documents/products/data-sheet/modules/unbuffered_dimm/ddr5/ddr5_udimm_core.pdf?rev=fdbd9476506c4e019360a5e402827caa https://media-www.micron.com/-/media/client/global/documents... Looking a bit closer I think you are right that it's 12V but only for Registered DIMMs.
deelowe 4y ago
There's been talk of eliminating sockets for years. Something has got to give.
- zeroth32 4y ago
  more compact chips will have a higher failure rate. Not a great idea for servers.
  - to11mtm 4y ago
    There's a fun curve on this to be sure. If I had to guess, Servers would not go any further than some sort of memory-backplane where the memory for multiple channels was integrated onto a single PCB. Even then, IIRC hot-swapping of memory modules is a thing for some servers, so that will have to be handled somehow.
- jnwatson 4y ago
  You can still actively cool socketed RAM.
  - deelowe 4y ago
    Sort of. Trace length is already a nightmare.
dcow 4y ago
> cooling has basically just been orientating the DIMMs to line up with airflow Isn't that server cooling in a nutshell? Ram high volumes of airflow through the chassis with stupidly loud fans and hope the parts stay cool?
- dodobirdlord 4y ago
  You still need to conduct the heat away from the sources to a radiator of some sort, since cooling is proportional to surface area and it’s much easier to increase surface area by adding fins than by increasing airflow. You can only speed up the air to a certain point, past which better cooling becomes a matter of shaping the components for more contact with the air.
  - kllrnohj 4y ago
    Sure but the airflow over DIMMs in a server chassis is already vastly more cooling than RAM gets in any consumer application other than GDDR on GPUs.
    - Ekaros 4y ago
      Yeah, kinda weird that on ATX RAM is placed in way that is perpendicular to usual CPU cooling or even general air flow. The top mounted fans do change this, but I don't think those are very common.
      - sbierwagen 4y ago
        Makes it easier to keep all the traces the same length: https://electronics.stackexchange.com/questions/74789/purpose-of-wave-shaped-pcb-traces https://electronics.stackexchange.com/questions/74789/purpos...
        kllrnohj 4y ago
        Although the entire socket can be rotated 90* for even better traces, which is what the EVGA Kingpin motherboards do ( https://www.evga.com/articles/01543/EVGA-Z690-DARK-KINGPIN/ https://www.evga.com/articles/01543/EVGA-Z690-DARK-KINGPIN/ )

AshamedCaptain 4y ago

I am not sure how much this is related to external cooling versus actually internal thermal dissipation. DDR JEDEC standards have actually decreased power consumption on every generation.

__alexs 4y ago
They have reduced voltage but power consumption per sq-mm has gone up with increased densities. Many people run DRAM at above JEDEC speeds which usually requires higher voltages too. Peak power consumption of DDR4 is around 375mW/GB @ 1.2V, DDR5 drops this about 10% but also increases the maximum density of a DIMM by 8x to 512GB which is like, 150W for a single DIMM.
- formerly_proven 4y ago
  There are only three (tiny) 12 V power pins on a DDR5 module, neither that nor the form factor allows for dissipating anywhere close to 150 W. The teased 512 GB Samsung module doesn't even have a heatspreader.
  - __alexs 4y ago
    VIN_BULK is 5V with a max current of 2A but every data pin provides current that is used on the DIMM in some respect.
    - formerly_proven 4y ago
      I thought they moved to 12 V, thanks for the correction!
      - __alexs 4y ago
        I'm just going by the Micron UDIMM data sheets. https://media-www.micron.com/-/media/client/global/documents/products/data-sheet/modules/unbuffered_dimm/ddr5/ddr5_udimm_core.pdf?rev=fdbd9476506c4e019360a5e402827caa https://media-www.micron.com/-/media/client/global/documents... Looking a bit closer I think you are right that it's 12V but only for Registered DIMMs.

deelowe 4y ago

There's been talk of eliminating sockets for years. Something has got to give.

zeroth32 4y ago
more compact chips will have a higher failure rate. Not a great idea for servers.
- to11mtm 4y ago
  There's a fun curve on this to be sure. If I had to guess, Servers would not go any further than some sort of memory-backplane where the memory for multiple channels was integrated onto a single PCB. Even then, IIRC hot-swapping of memory modules is a thing for some servers, so that will have to be handled somehow.
jnwatson 4y ago
You can still actively cool socketed RAM.
- deelowe 4y ago
  Sort of. Trace length is already a nightmare.

dcow 4y ago

> cooling has basically just been orientating the DIMMs to line up with airflow Isn't that server cooling in a nutshell? Ram high volumes of airflow through the chassis with stupidly loud fans and hope the parts stay cool?

dodobirdlord 4y ago
You still need to conduct the heat away from the sources to a radiator of some sort, since cooling is proportional to surface area and it’s much easier to increase surface area by adding fins than by increasing airflow. You can only speed up the air to a certain point, past which better cooling becomes a matter of shaping the components for more contact with the air.
- kllrnohj 4y ago
  Sure but the airflow over DIMMs in a server chassis is already vastly more cooling than RAM gets in any consumer application other than GDDR on GPUs.
  - Ekaros 4y ago
    Yeah, kinda weird that on ATX RAM is placed in way that is perpendicular to usual CPU cooling or even general air flow. The top mounted fans do change this, but I don't think those are very common.
    - sbierwagen 4y ago
      Makes it easier to keep all the traces the same length: https://electronics.stackexchange.com/questions/74789/purpose-of-wave-shaped-pcb-traces https://electronics.stackexchange.com/questions/74789/purpos...
      - kllrnohj 4y ago
        Although the entire socket can be rotated 90* for even better traces, which is what the EVGA Kingpin motherboards do ( https://www.evga.com/articles/01543/EVGA-Z690-DARK-KINGPIN/ https://www.evga.com/articles/01543/EVGA-Z690-DARK-KINGPIN/ )