Monero Mining Performance

[Patrick]

[2017-06-21 22:55:29] accepted: 484/484 (100.00%), 3526.99 H/s at diff 10000 (yay!!!)
[2017-06-21 22:55:33] accepted: 485/485 (100.00%), 3530.57 H/s at diff 10000 (yay!!!)
[2017-06-21 22:55:37] accepted: 486/486 (100.00%), 3530.49 H/s at diff 10000 (yay!!!)
[2017-06-21 22:55:39] accepted: 487/487 (100.00%), 3530.20 H/s at diff 10000 (yay!!!)
[2017-06-21 22:55:44] accepted: 488/488 (100.00%), 3530.49 H/s at diff 10000 (yay!!!)
[2017-06-21 22:55:44] accepted: 489/489 (100.00%), 3529.14 H/s at diff 10000 (yay!!!)
[2017-06-21 22:55:45] accepted: 490/490 (100.00%), 3524.09 H/s at diff 10000 (yay!!!)
[2017-06-21 22:55:52] accepted: 491/491 (100.00%), 3530.97 H/s at diff 10000 (yay!!!)
[2017-06-21 22:55:52] accepted: 492/492 (100.00%), 3529.75 H/s at diff 10000 (yay!!!)
[2017-06-21 22:55:54] accepted: 493/493 (100.00%), 3531.22 H/s at diff 10000 (yay!!!)
[2017-06-21 22:55:55] accepted: 494/494 (100.00%), 3530.36 H/s at diff 10000 (yay!!!)

 

[blakwolf]

holy moly...thats a lotta processing power...

 

[Serious Master]

Taking AWS monero mining notes:

c4.8xlarge
c4.8xlarge = 734 H/s w/ 36 threads
c4.8xlarge = 959 H/s w/ 24 threads
c4.8xlarge = 979 H/s w/ 25 threads
c4.8xlarge = 941 H/s w/ 26 threads

Verdict: L3 cache via lscpu is the best. Not profitable even with spot instance pricing.

r4.4xlarge
r4.4xlarge = 530 H/s w/ 15 threads
r4.4xlarge = 560 H/s w/ 16 threads
r4.4xlarge = 560 H/s w/ 17 threads
r4.4xlarge = 558 H/s w/ 18 threads
r4.4xlarge = 525 H/s w/ 24 threads

r4.2xlarge
r4.2xlarge = 287 H/s w/ 8 threads
r4.2xlarge = 300 H/s w/ 9 threads
r4.2xlarge = 290 H/s w/ 10 threads
r4.2xlarge = 284 H/s w/ 16 threads

The numbers are jumping quite a bit on all of the AWS instances. Likely due to resource contention. All instances launched on E5-2686 V4 machines.

Am getting average 1020 H/s on c4.8xlarge with 23 threads

 

[Patrick]

@Serious Master will check it out. AWS does have some variance with contention issues.

 

[tgsz]

4x Opteron 6334: 1600 H/s
4x Opteron 6234: 1500 H/s


System (4x + 8 sticks of ddr3 and dual 10gbe nic) pulling 580W from the wall with 1400W platinum power supply. 75% cpu core usage (36/48). Could probably get down to ~515W usage by making it run on 32 cores only.



------

Note you need to use 8 threads per physical cpu since they have 16MB L3 per cpu. If you run the miner which auto-detects L3 and threads, i twill only run 12 threads - so you have to explicitly run 3 instances of the miner with no flags, or force 1 instance to use -t 32. Running 3 instances will run 36 cores, so not ideal, but since I'm using the STH container - without hacking the entry script its difficult (@Patrick maybe you can make this an env variable?)


------
4x 6334:

(per instance, 3 of these instances running):
[2017-06-26 19:04:36] accepted: 13473/13525 (99.62%), 561.12 H/s at diff 5440 (yay!!!)
[2017-06-26 19:04:41] accepted: 13474/13526 (99.62%), 540.75 H/s at diff 5440 (yay!!!)
[2017-06-26 19:04:46] accepted: 13475/13527 (99.62%), 557.22 H/s at diff 5440 (yay!!!)
[2017-06-26 19:04:50] accepted: 13476/13528 (99.62%), 585.81 H/s at diff 5440 (yay!!!)
[2017-06-26 19:05:00] accepted: 13477/13529 (99.62%), 546.43 H/s at diff 5440 (yay!!!)
[2017-06-26 19:05:09] accepted: 13478/13530 (99.62%), 563.38 H/s at diff 5440 (yay!!!)
[2017-06-26 19:05:27] accepted: 13479/13531 (99.62%), 553.51 H/s at diff 5440 (yay!!!)
[2017-06-26 19:05:27] accepted: 13480/13532 (99.62%), 540.64 H/s at diff 5440 (yay!!!)
[2017-06-26 19:05:33] accepted: 13481/13533 (99.62%), 546.56 H/s at diff 5440 (yay!!!)

*-cpu:3
description: CPU
product: AMD Opteron(tm) Processor 6344
vendor: Advanced Micro Devices [AMD]
physical id: 10
bus info: cpu@3
version: AMD Opteron(tm) Processor 6344
serial: To Be Filled By O.E.M.
slot: CPU 4
size: 2600MHz
capacity: 2600MHz
width: 64 bits
clock: 200MHz
capabilities: x86-64 fpu fpu_exception wp 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 constant_tsc rep_good nopl nonstop_tsc extd_apicid amd_dcm aperfmperf pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 popcnt aes xsave avx f16c lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs xop skinit wdt lwp fma4 tce nodeid_msr tbm topoext perfctr_core perfctr_nb arat cpb hw_pstate npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold vmmcall bmi1 cpufreq
configuration: cores=12 enabledcores=12 threads=12

4x 6234:

(per instance, 3 of these instances running):

[2017-06-26 19:04:38] accepted: 13507/13540 (99.76%), 508.88 H/s at diff 4900.01 (yay!!!)
[2017-06-26 19:04:53] accepted: 13508/13541 (99.76%), 513.71 H/s at diff 4900.01 (yay!!!)
[2017-06-26 19:05:05] accepted: 13509/13542 (99.76%), 504.15 H/s at diff 4900.01 (yay!!!)
[2017-06-26 19:05:09] accepted: 13510/13543 (99.76%), 487.71 H/s at diff 4900.01 (yay!!!)
[2017-06-26 19:05:09] accepted: 13511/13544 (99.76%), 488.88 H/s at diff 4900.01 (yay!!!)

*-cpu:3
description: CPU
product: AMD Opteron(TM) Processor 6234
vendor: Advanced Micro Devices [AMD]
physical id: 10
bus info: cpu@3
version: AMD Opteron(TM) Processor 6234
serial: To Be Filled By O.E.M.
slot: CPU 4
size: 2400MHz
capacity: 2400MHz
width: 64 bits
clock: 200MHz
capabilities: x86-64 fpu fpu_exception wp 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 constant_tsc rep_good nopl nonstop_tsc 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 arat cpb hw_pstate npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold vmmcall cpufreq
configuration: cores=12 enabledcores=12 threads=12

 

[tgsz]

4x E7-8837 (8c, 2.66ghz, 24MB L3)

1110 H/s

Draws 630W at the wall in an R810. Uses 100% cpu.

-26 20:50:05] accepted: 78/78 (100.00%), 1125.81 H/s at diff 1063 (yay!!!)
-26 20:50:05] accepted: 79/79 (100.00%), 1134.75 H/s at diff 1063 (yay!!!)
-26 20:50:06] accepted: 80/80 (100.00%), 1119.52 H/s at diff 1063 (yay!!!)
-26 20:50:06] accepted: 81/81 (100.00%), 1110.08 H/s at diff 1063 (yay!!!)
-26 20:50:06] accepted: 82/82 (100.00%), 1090.50 H/s at diff 1063 (yay!!!)
-26 20:50:07] accepted: 83/83 (100.00%), 1113.54 H/s at diff 1063 (yay!!!)
-26 20:50:08] accepted: 84/84 (100.00%), 1116.51 H/s at diff 1063 (yay!!!)


*-cpu:3
description: CPU
product: Xeon
vendor: Intel Corp.
vendor_id: GenuineIntel
physical id: 403
bus info: cpu@3
version: Intel(R) Xeon(R) CPU E7- 8837 @ 2.67GHz
slot: CPU4
size: 2666MHz
capacity: 3600MHz
width: 64 bits
clock: 2105MHz
capabilities: x86-64 fpu fpu_exception wp vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic popcnt aes lahf_lm ida arat dtherm tpr_shadow vnmi flexpriority ept vpid
configuration: cores=8 enabledcores=8 threads=8

 

[Marsh]

Draws 630W at the wall in an R810.

I hope that you do not pay for the electricity , otherwise you will loose money.
According to my calculation, it is like 1.76 H/s per watt.
At 10 cent per KW/h , you will make $35 per month.
At 18 cent per KW/h, you are negative $3 per month.

 

[tgsz]

I hope that you do not pay for the electricity , otherwise you will loose money.
According to my calculation, it is like 1.76 H/s per watt.
At 10 cent per KW/h , you will make $35 per month.
At 18 cent per KW/h, you are negative $3 per month.

Typically in colocation you pay for the capacity not the consumption. So in this case the power is already paid for.

Additionally electricity here is $.06/kWh (USD) so it's still profitable.

The AMD system has a bit better profitability.

 

[Marsh]

Where are you located?

If my electricity rate is 6 cent, I'll turn on all my E5-26xx v1 nodes, bring down the power transformer of my street.
Mine , Mine , Mine all day and all night like Snow White and 7 dwarfs.

I looked up the Song Lyrics: Snow White and the 7 Dwarfs

Heigh-Ho
We dig dig dig dig dig dig dig in our mine the whole day through
To dig dig dig dig dig dig dig is what we really like to do
It ain't no trick to get rich quick
If you dig dig dig with a shovel or a pick
In a mine! In a mine! In a mine! In a mine!
Where a million diamonds shine!

We dig dig dig dig dig dig dig from early morn till night
We dig dig dig dig dig dig dig up everything in sight
We dig up diamonds by the score
A thousand rubies, sometimes more
But we don't know what we dig 'em for
We dig dig dig a-dig dig

Heigh-ho, Heigh-ho
Heigh-ho, Heigh-ho
Heigh-ho

 

[tgsz]

It's actually a bit less than that, depending on the exchange.

 

[Minyó András]

Hello,

What did i wrong?
I have Ryzen 5 1600 and i dig monero with ~260H/s. I have a windows 10 and I use MinerGate v.6.8.
@Patrick , how did you do 363H/s with Ryzen 5 1600?

 

[Marsh]

windows 10 and I use MinerGate v.6.8

Suggestion:
1) using another miner program such as xmr-stak-cpu , mine with another pool. I use minexmr.com.
2) Change to Linux OS.
I am pretty sure you will see hashrate increase with these changes.

 

[pyro_]

Try using something other than minergate. I noticed i got higher performance when i switched over from it to something like XMR-Stak.

Also how many threads to you have running on your 1600?

 

[OMinerO]

Hello,

First would like to thanks this website and all the contributors of this superb topic! i didnt read all the pages yet but it's the only place I found on google with very useful informations.

I have a sleeping dual Xeon 2670 v1 and wanted to test CPU mining.

Configuration is:

2x 2670
64Gb ram
SSD
XMR Stak CPU
minerXMR - 200000 difficulty - port 3333
Windows 10


I got like 814 H/s. I think my config file isnt really good since I had a lot of problem making it but went from 200 to 814 and thanks to your forum it seems 814 is normal for a dual 2670.

But i wonder why i only use 50% of the CPU Power and half of the core? On task manager and other application i can see 50-55% CPU use. Is it normal?

Thanks!

 

[pyro_]

With it only having 20mb cache and each thread needing 2mb you will only use max 10 threads per CPU

 

[Patrick]

Try Dwarfpool and use Docker in Linux and you will be set.

 

[OMinerO]

Thank you for your replies.

I changed the config so i use 20 threads (10 per CPU) made it to use more different physical core to have a 50% on them.
I moved to dwarf and now with 72% CPU use I got: 915 H/s

 

[Minyó András]

Suggestion:
1) using another miner program such as xmr-stak-cpu , mine with another pool. I use minexmr.com.
2) Change to Linux OS.
I am pretty sure you will see hashrate increase with these changes.

Thank you so much. I changed the minergate to xmr-stak-cpu and only this change increase 25% hashrate, now with ryzen 5 1600 hashrate is 366H/s.

Okay, next time i will see the minerxmr pool and dwarf pool. How i know which is better? Hashrate or Mined xmr per day?

 

[Marsh]

Hashrate or Mined xmr per day

The only way to know is to try both, I keep a log of xmr mined , average hashrate , at 8am , and 8pm to compare .
example:
Average Total Hash Rate: (24h) 16.56 KH/s (12h) 16.66 KH/s (1h) 17.13 KH/s (10m) 23.28 KH/s

There is also a element of "luck" involved. so find a pool that you are comfortable with or feeling lucky

Look for a pool are transparent with their operation.
Take a look at mineXMR.com - Monero mining pool , it show you the "luck" factor, and other stats

 

[Klee]

Gigabyte Aorus GTX 1080 Ti Xtreme Edition....IT IS A BEAST!!

Playing around with XMR-STAK-NVIDIA in Windows 10, could not get it running in Linux as it complained about the version off gcc was incompatible with Cuda 8 in Linux and to use clang and that errored out also, and [playing around with MSI Afterburner.

Afterburner settings:
Power Limit %150
Temp Limit 90
Core Clock +100
Memory Clock +200
Fan Speed %100

First thing I noticed was that the fans at %100 were barely louder than the system fans.

I just spent maybe 5 minutes playing around with afterburners settings and they are not even close to being optimized, just a quick bump.

2100 Mhz gpu speed
5315 Mhz memory speed

AND 38 c temp...... It literally runs at half the temp of my rx480's.

Results 792 H/s.

Not even tweaked to the ragged edge at all.

I need to get it working under Linux and there seems to be a patch for Cuda 8 that came out a couple of days ago.