EU Supermicro 4 Node Server 2022TG 8x AMD Opteron 6275 16 Core CPU G34 @£199 Collection Only

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

anoother

Member
Dec 2, 2016
135
24
18
35
These look super interesting, BUT:

- I can find no mention of them on the SM site, or anywhere in fact
- The heatsinks look like Intel heatsinks... Not even sure if they're big enough to cover a G34 CPU.

They have SGI/Rackable stickers, so likely custom OEM solution by SM?
 

michathe

Active Member
Feb 25, 2016
128
44
28
37
These look super interesting, BUT:

- I can find no mention of them on the SM site, or anywhere in fact
- The heatsinks look like Intel heatsinks... Not even sure if they're big enough to cover a G34 CPU.

They have SGI/Rackable stickers, so likely custom OEM solution by SM?
I thougt these where LGA1366 at first but Supermicro makes a Motherboard that seems do use intel Heatsinks on the G34 socket.

Super Micro Computer, Inc. - Aplus Products | Motherboards | H8DGT-HLIBQF
 

anoother

Member
Dec 2, 2016
135
24
18
35
I thougt these where LGA1366 at first but Supermicro makes a Motherboard that seems do use intel Heatsinks on the G34 socket.

Super Micro Computer, Inc. - Aplus Products | Motherboards | H8DGT-HLIBQF
Yup, was just looking into it. That does look like the board, and this looks like the system: Super Micro Computer, Inc. - Products | A+ Servers | 2U | 2022TG-HLIBQRF

You can also see in the seller images how the socket almost meets the edge of the heatsink on one end, so it definitely looks G34 sized.

I wonder how much the FLOPS limitiation impacts performance; I guess these were designed to be used for network pushing packets around at high-speed, where it probably doesn't make too much difference.

I don't need a loud, power-hungry, relatively-low-performance cluster, but I am mighty tempted...
 

Mishka

Active Member
Apr 30, 2017
100
33
28
London, UK
I wonder how much the FLOPS limitiation impacts performance; I guess these were designed to be used for network pushing packets around at high-speed, where it probably doesn't make too much difference.
Wonder if you could swap out the CPU for a 6272 or a 6276, both fairly common and cheap (£20-30 per CPU) although they are getting a bit rare it seems for the cheap ones
 

Iaroslav

Active Member
Aug 23, 2017
117
28
28
38
Kyiv
Ah, yes... They were going too fast. Maybe later we will see them again for a higher price.
 

linuxmanbg

Member
Jun 23, 2016
68
23
8
51
Hello,
a day before weekend my 2 SuperMicro servers arrived, but now I have time to publish some pics.

out2smal2.jpg outt.jpg out3smal.jpgout4smal.jpg out5smal.jpg
the motherbord is found on SM site it is: H8DGT-HLIBQF-SG007
BIOS was 3.5 upgraded to 3.5c
IPMI FW was 2.1x upgraded to 3.15 .
Select in BIOS for all working nodes (currently 2) Energy saving and FAN is not noisy (but they become noisy if 80% of cpus are busy), two nodes mining monero (soft. xmr-stak-cpu ) 77-82 dB .
 
Last edited:

Iaroslav

Active Member
Aug 23, 2017
117
28
28
38
Kyiv
Hello,
a day before weekend my 2 SuperMicro servers arrived, but now I have time to publish some pics.

View attachment 7380 View attachment 7381 View attachment 7382View attachment 7383 View attachment 7384
the motherbord is found on SM site it is: H8DGT-HLIBQF-SG007
BIOS was 3.5 upgraded to 3.5c
IPMI FW was 2.1x upgraded to 3.15 .
Select in BIOS for all working nodes (currently 2) Energy saving and FAN is not noisy (but they become noisy if 80% of cpus is busy), two nodes mining monero (soft. xmr-stak-cpu ) 77-82 dB .
Great, mine is still on its way.
So, what is your actual hashrate per node? As far as I see it's ok with a non-ECC memory?
 

linuxmanbg

Member
Jun 23, 2016
68
23
8
51
Great, mine is still on its way.
So, what is your actual hashrate per node? As far as I see it's ok with a non-ECC memory?
Fedora 25 ( 4.13.16-100.fc25.x86_64) 22 threads :
Totals: 790.1 791.4 783.8 H/s
Highest: 839.0 H/s
sometimes: 914 H/s

Fedora 26 (4.14.8-200.fc26.x86_64) 22 threads:
Totals: 940.9 941.6 942.0 H/s
Highest: 954.3 H/s
2 nodes consume ~642W measured with watmeter.
Accidently put non buffered memoty 2x4GB and it worked (officially supported)!
lscpu:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 32
On-line CPU(s) list: 0-31
Thread(s) per core: 2
Core(s) per socket: 8
Socket(s): 2
NUMA node(s): 4
Vendor ID: AuthenticAMD
CPU family: 21
Model: 1
Model name: AMD Opteron(tm) Processor 6275
Stepping: 2
CPU MHz: 2300.000
CPU max MHz: 2300,0000
CPU min MHz: 1400,0000
BogoMIPS: 4600.09
Virtualization: AMD-V
L1d cache: 16K
L1i cache: 64K
L2 cache: 2048K
L3 cache: 6144K
NUMA node0 CPU(s): 0-7
NUMA node1 CPU(s): 8-15
NUMA node2 CPU(s): 16-23
NUMA node3 CPU(s): 24-31
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid amd_dcm aperfmperf pni pclmulqdq monitor ssse3 cx16 sse4_1 sse4_2 popcnt aes xsave avx lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs xop skinit wdt lwp fma4 nodeid_msr topoext perfctr_core perfctr_nb cpb hw_pstate vmmcall arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold

power info:
cpupower frequency-info
analyzing CPU 0:
driver: acpi-cpufreq
CPUs which run at the same hardware frequency: 0
CPUs which need to have their frequency coordinated by software: 0
maximum transition latency: 5.0 us
hardware limits: 1.40 GHz - 2.30 GHz
available frequency steps: 2.30 GHz, 1.40 GHz
available cpufreq governors: conservative userspace powersave ondemand performance schedutil
current policy: frequency should be within 1.40 GHz and 2.30 GHz.
The governor "ondemand" may decide which speed to use
within this range.
current CPU frequency: 1.40 GHz (asserted by call to hardware)
boost state support:
Supported: yes
Active: yes
Boost States: 2
Total States: 7
Pstate-Pb0: 3200MHz (boost state)
Pstate-Pb1: 2600MHz (boost state)
Pstate-P0: 2300MHz
Pstate-P1: 2300MHz
Pstate-P2: 2300MHz
Pstate-P3: 2300MHz
Pstate-P4: 1400MHz

turbostat
turbostat version 17.06.23 - Len Brown <lenb@kernel.org>
CPUID(0): AuthenticAMD 13 CPUID levels; family:model:stepping 0xf:1:2 (15:1:2)
CPUID(1): SSE3 MONITOR - - - TSC MSR - -
CPUID(6): APERF, No-TURBO, No-DTS, No-PTM, No-HWP, No-HWPnotify, No-HWPwindow, No-HWPepp, No-HWPpkg, No-EPB
CPUID(7): No-SGX
cpu3: POLL: CPUIDLE CORE POLL IDLE
cpu3: C1: <null>
cpu3: C2: ACPI IOPORT 0x815
cpu3: cpufreq driver: acpi-cpufreq
cpu3: cpufreq governor: ondemand
cpufreq boost: 1
Package Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz IRQ C1 C2 C1% C2%
- - - 894 68.76 1300 1150 66047 13 279 0.00 15.62
0 0 8 2295 88.23 2602 2301 5259 0 1 0.00 11.78
0 1 9 1585 60.92 2602 2301 3661 2 22 0.00 39.11
0 2 10 1397 53.69 2602 2302 3202 0 4 0.00 46.35
0 3 11 2583 99.23 2603 2302 5922 0 13 0.00 0.77
0 4 12 484 18.61 2601 2301 1130 1 6 0.00 81.41
 
  • Like
Reactions: Iaroslav

BlackArchon

Member
Jun 23, 2016
71
10
8
Thank you, linuxmanbg!
These turbo steps looks exactly like the 6276, the core configuration too. Maybe the 6275 has a lower TDP than the 6276's 115 W, but I don't know how to get this value within Linux. Can someone help?