BUILD Supermicro MBD-X9SRH-7F-O CPU Heatsink recommendation

[Binary Buddha]

I need some help finding a suitable heatsink for my new motherboard; Supermicro MBD-X9SRH-7F-O.

I picked up a Corsair Hydro Series H80i v2 Extreme Performance Liquid CPU Cooler and the bracket doesn't seem to line up.

Anybody got this motherboard? What heatsink do you have? Keep in mind I'm trying to stick to liquid cooling because I'm in Hawaii and the server is a private server hiding under my bed.

 

[Stux]

That is an LGA 2011 Narrow ILM socket/bracket. It's narrower than a regular socket. You need a cooler which supports Narrow ILM sockets.

https://www.servethehome.com/narrow-ilm-square-ilm-lga2011-heatsink-differences/

I personally use this cooler:
http://noctua.at/en/products/cpu-cooler-workstation-server/nh-u9dx-i4

Very quiet. Especially when operating at low speed.

 

[Spearfoot]

I need some help finding a suitable heatsink for my new motherboard; Supermicro MBD-X9SRH-7F-O.

I picked up a Corsair Hydro Series H80i v2 Extreme Performance Liquid CPU Cooler and the bracket doesn't seem to line up.

Anybody got this motherboard? What heatsink do you have? Keep in mind I'm trying to stick to liquid cooling because I'm in Hawaii and the server is a private server hiding under my bed.

Supermicro makes two CPU coolers that will work:

http://store.supermicro.com/heatsin...active=148&heat_sink_supported_cpu_socket=174

I use two of the SNK-P0050AP4 CPU coolers in my X9DRi-based system. The build quality/fit-and-finish is excellent.

 

[Binary Buddha]

Picking up the Noctua NH-U9DXi4. Hopefully it works for me. I'm not that acquainted with the new heat pipe heatsinks.

 

[svtkobra7]

Picking up the Noctua NH-U9DXi4. Hopefully it works for me. I'm not that acquainted with the new heat pipe heatsinks.

Currently have 2x Noctua NH-U9DXi4, but ordered a pair of SuperMicro SNK-P048AP4 CPU fans which are in transit (one of the two @Spearfoot called out). You likely made the right choice, as it is quite quiet but I suspect when I line them up I'm going to get lower CPU temps with the Supermicro fan. I didn't notice any improvement moving from passive heatsink + airshroud to the Noctuas.

@Spearfoot @Stux, I'm quite eager to do an apples:apples comparison since I'll have both fans. Your thoughts on how I can create a controlled environment for testing, I. E. CPU @ idle, 50%, 75%, etc. for purposes of measuring?

Picking up the Noctua NH-U9DXi4. Hopefully it works for me. I'm not that acquainted with the new heat pipe heatsinks.

Sent from my HTC6545LVW using Tapatalk

 

[Spearfoot]

@Spearfoot @Stux, I'm quite eager to do an apples:apples comparison since I'll have both fans. Your thoughts on how I can create a controlled environment for testing, I. E. CPU @ idle, 50%, 75%, etc. for purposes of measuring?

I don't know of any way to reliably run a system at a specific CPU utilization, but you could easily obtain the endpoint temps using a bootable DOS USB drive and the Ultimate Boot CD:

• Boot DOS. The CPUs will effectively be idling because the only running process will be the DOS shell. This will give you the 'Idle CPU' temps.
  
• Boot the Ultimate Boot CD and run the Mersenne Prime Test; this should run the CPUs at pretty close to 100%. This will give you the 'Max CPU utilization' temps.

In both cases, record the temps with the fans running at the various settings: Full, Standard, Optimal, Heavy IO.

Warning: Don't run the fans at low speeds very long when you're running the Mersenne Prime Test, 'cause it really heats up the CPUs!

 

[svtkobra7]

@Binary Buddha: Sorry to threadjack. ;)

I don't know of any way to reliably run a system at a specific CPU utilization, but you could easily obtain the endpoint temps using a bootable DOS USB drive and the Ultimate Boot CD:

• Boot DOS. The CPUs will effectively be idling because the only running process will be the DOS shell. This will give you the 'Idle CPU' temps.
  
• Boot the Ultimate Boot CD and run the Mersenne Prime Test; this should run the CPUs at pretty close to 100%. This will give you the 'Max CPU utilization' temps.

• Thanks for the recommendations. I knew there was some kind of load tester on the boot CD, but haven't used it in a while and couldn't remember.
• I searched long and hard for a comparison of the two fans, but couldn't find one.
• A "controlled" comparison is selfish in that it helps me select the optimal fan for my circumstances, but hopefully empirical analysis and public presentation of such will benefit the community, specifically anyone that is considering the two different fans. (I will create a separate thread for presentation of that analysis)

In both cases, record the temps with the fans running at the various settings: Full, Standard, Optimal, Heavy IO.

• As if this recommendation need be made. ;) (I was planning on such)
  
• You do recall my last empirical analysis, eh? ;)

Warning: Don't run the fans at low speeds very long when you're running the Mersenne Prime Test, 'cause it really heats up the CPUs!

• Bah - they are indestructible Xeons designed to withstand high temps and 24/7 use. They will throttle anyway at Tmax, no? <= [Famous last words before I start scouring the 'bay for 2x E5-2670 v2] LOL :)
• Unfamiliar with this Mersenne Prime, is he Optimus' evil brother? [why do I have such a bad sense of humor???]
  
• Seriously though, your advice is noted and thank you.