AMD Epyc + Synology build

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

n0xlf

New Member
Nov 5, 2018
9
5
3
It was time to replace my aging Dell Poweredge 2900 and Synology 1512+, so I decided on a major upgrade and learned a few things along the way that I thought I'd pass on.

Specs:

Supermicro AS-1113S-WN10RT (1113S-WN10RT | 1U | A+ Servers | Products - Super Micro Computer, Inc.)
  • AMD Epyc 7551P 32 core CPU
  • 8x 16GB Samsung ECC Registered single rank c-die @ 2666MHz
  • Intel DC P4610 1.6TB SSD
Synology Rackstation RS3618xs (RackStation RS3618xs | Synology Inc.)
  • 6x Seagate ST12000NM0007 (12TB Exos X12)
  • RKS1317 rail kit
  • Synology E10G18-T1 10GbE PCIe expansion card (copper)
  • Crucial CT16G4WFD8266 16GB DDR4 2666MHz UDIMM
Some of this is obvious, but I wanted to include it for anyone considering such a build. Might save you some time searching!

The memory, in particular, required some research. There is b-die and c-die memory from Samsung that is very similar but b-die is a 20nm process and c-die is 10nm.

AMD Document about the Epyc and memory: https://developer.amd.com/wp-content/resources/56301_1.0.pdf
B-Die specs: https://www.samsung.com/semiconduct..._8Gb_B_die_Registered_DIMM_Rev1.91_May.17.pdf
C-Die specs: https://www.samsung.com/semiconduct...4_8Gb_C_die_Registered_DIMM_Rev1.3_Feb.18.pdf
C-Die article: Samsung Starts Production of 8 Gb DDR4-3600 ICs Using 2nd Gen 10nm-Class Tech
Supermicro B-Die: https://store.supermicro.com/16gb-ddr4-2666-mem-dr416l-sl02-er26.html
Supermicro C-Die: https://store.supermicro.com/16gb-ddr4-2666-mem-dr416l-sl04-er26.html

B-Die is quite the rage for good performance when searching, but there is no reason why c-die, being a 10nm process, shouldn't outperform it in every way. The spec sheets, if I'm reading them correctly, seem to show a slight edge as well. If anyone knows anything different about this, please chime in.

I also found that the info regarding single rank vs dual rank when trying to run @2666 is limited. The AMD document above touches on it a bit, showing that single rank or LRDIMMs are the only ones running at 2666. Running dual rank at 2666 could perform better, but the data is mixed with different motherboards and BIOSes behaving differently when that is attempted. Overall, I found that running no less than 8 DIMMs and running at 2666 were both key to maximum performance, also partially due to the AMD infinity fabric operating at memory speed. Be aware of how the Epyc handles memory and creates NUMA nodes. This could affect performance depending on your application. It's one of the few downsides of these CPUs vs. Intel.

I only went with 6 of the 12TB drives since I don't need more storage at the moment and thus might as well lower the current running cost, future drive cost, and spread out the drive failures a bit. The Synology server will primarily be iSCSI for the Supermicro server.

The Intel SSD is quite a beast (Intel® SSD DC P4610 Series (1.6TB, 2.5in PCIe* 3.1 x4, 3D2, TLC)) both in speed and lifetime endurance.

Waiting for the Mikrotik CRS312-4C+8XG to come out, and that will allow me to do copper 10gig to the Supermicro (it has 2 10gig ports) and Synology.

The Synology will be here next week - The Supermicro hasn't shipped. I've been dealing with atacom.com, who have been great and are local to Supermicro, so that's handy. Getting distributors to respond is difficult and changing things around/asking questions is not a fast process so don't expect immediate satisfaction if you go down a similar road.

It's pretty awesome to see a tiny SSD replacing an entire 15k array that I have now, and stomp all over it in performance! Looking forward to getting all of it!
 
Last edited:
  • Like
Reactions: Patrick

Evan

Well-Known Member
Jan 6, 2016
3,346
598
113
Unless you want the potential power saving of C-die then I would not be worried at all about using B-die, it’s so far much more common and available.
One reason memory has stayed until just recently at the larger size production is also reliability, I don’t know the details but the reliability and also longevity at 20nm is better, of course that is news from a year+ ago so may not be an issue now.

I assume major and first use for C-die will be LPDDD4 for portable devices.

Off topic maybe for TR (and I know little about TR specifics) I noticed some intel systems can’t use single rank at all (and some also not LRDIMM) I would have said 2Rx4 was the almost 100% assured choice to work in intel systems.
 

n0xlf

New Member
Nov 5, 2018
9
5
3
Looks like c-die is happening! Either way would have been fine with me, but can't complain.
 

n0xlf

New Member
Nov 5, 2018
9
5
3
The better matching memory to the modules/Samsung chips that Synology uses is the Samsung M391A2K43BB1-CTD.

It's very hard to find, so I ended up getting one off of EBay.
 

n0xlf

New Member
Nov 5, 2018
9
5
3
Finally got the Supermicro!

Old vs New: Supermicro AS -1113S-WN10RT vs Dell Inc. PowerEdge 2900 - Geekbench Browser

Code:
%Cpu0  :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu1  :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu2  :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu3  :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu4  :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu5  :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu6  :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu7  :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu8  :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu9  :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu10 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu11 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu12 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu13 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu14 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu15 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu16 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu17 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu18 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu19 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu20 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu21 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu22 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu23 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu24 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu25 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu26 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu27 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu28 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu29 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu30 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu31 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu32 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu33 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu34 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu35 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu36 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu37 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu38 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu39 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu40 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu41 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu42 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu43 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu44 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu45 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu46 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu47 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu48 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu49 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu50 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu51 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu52 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu53 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu54 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu55 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu56 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu57 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu58 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu59 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu60 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu61 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu62 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu63 :  0.0 us,  0.0 sy,  0.0 ni,100.0 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
MiB Mem : 128799.4 total, 127404.2 free,    853.2 used,    542.1 buff/cache
MiB Swap:   4096.0 total,   4096.0 free,      0.0 used. 127157.4 avail Mem
Idle power is 86 watts. I got it to 286 under 100% load Supermicro got it to 304 during burn in (logged in the BMC). It's incredibly silent idle and incredible loud at 100% load, as would be expected for a 1U with tiny fans.

It has an onboard microSD card slot, which I thought would be useful for recovery image(s)/data, but it doesn't seem to be exposed to anything as anything. It's not a drive on a controller and it's not in the BMC so I'll have to ask Supermicro about that one. I would think that it's purpose is primary to boot something like ESXi.

Just having fun doing mostly pointless stuff at the moment :)
 

n0xlf

New Member
Nov 5, 2018
9
5
3
Response from Supermicro on the microSD card slot in case anyone ever tries this themselves:
For this matter, the MicroSD slot on this board does not support bootable device.
So if you have an OS in the MicroSD, by design, it will not be able to boot into the OS.