Topic: Building Cheap Miners : My "Secret" - page 24. (Read 60249 times)

Its kind of annoying having dual 1400W Gold class PSU in the server but no PCIe power connectors from the PSUs.
So there is breakout-boards that can solve this so I can use existing PSUs?

My server is Supermicro 2042-6RF.

Breakout boards...  All my HP server iron have GPUs running along with the CPUs.  Boxes like the r815 that don't have the onboard connectors to power them but could easily run GPUs with the addition of external PSU and risers.

Some servers do have PCIe slots and VGA power cables. My Supermicro server is a dual 6348 Opteron that has dual 1200 watt power supplies and has 6 PCIe slots and four 6-pin VGA power cables.

Expansion Slots
• Three (3) PCI-Express x16 Gen. 2
• One (1) PCI-Express x8 Gen. 2
• Two (2) PCI-Express x4 (in x8 slot) Gen. 2

I currently have three Nvidia GTX 750 in the three PCI-Express x16 Gen. 2 slots.

Aha, you are mining IntenceCoin :-)

Regarding filling empty PCIe slots I think these server power supplies does not have PCIe power connectors? So we are limited to cards with power from PCIe slot ony or using external power supplies? I have a Supermicro quad socket Opteron 6276 machine but CPU mining only on that one. I figured out that by using external power supply and powered risers it may be possible using GPUs too but right now I have a separate GPU mining rig.

Please show me where you can purchase these cards in bulk at a price that is not astronomical as I would love to fill my empty PCIe slots with them.  Some of us are on the path of putting whatever iron on the fire that is making money, even if it isn't exactly the most efficient or latest wiz bang contraption.  Even better, when / if () GPU prices get back to normal, we have tons of open slots ready to accept them with minimal config and downtime.  Until then, they at the minimum, more than cover the expense for electricity of the combined GPU/CPU mining activity and earn back their own cost, slowly but surely, even in the depressed market we are in compared to 3 months ago.

-=Slow Growth=-

Someone might see what I did there...

Until then, I snipe cards when and where I can if they are relatively close to what msrp was.

What's the advantage of building a $500 system doing 1600hs and then building 60x of them? That's still $30k. When you can get the same hash with just 2x 4xx/5xx cards or 1x vega and still have more slots to lower the system power overhead? Just spend your $30k on modern, faster, more efficient hardware...

Upon further review it appears that Sundownz is on the right track with 1x threads on every core. These Opterons behave differently than the Xeons that I have optimized before.

The 6300 series Opterons have a CC (Core Complex) that consists of 2 integer cores stuck together.

When I used:

1x thread on one of the two cores in the CC results in 65 H/s.
2x thread on one of the two cores in the CC results in 75 H/s.
1x threads on both of the two cores in the CC results in 82 H/s.

Power does go up with more cores running but not as much as I expected.

For some reason the hash rate does not drop much when the thread count exceeds the L2/L3 cache. When running all 12 cores the hash rate goes to 39/40 from 41.

I am now running this configuration on my dual 6348 Opteron system:

"cpu_threads_conf" :
   [
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 0  },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 2  },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 3  },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 4  },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 5  },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 6  },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 7  },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 8  },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 9  },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 10 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 11 },

        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 12 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 13 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 14 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 15 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 16 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 17 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 18 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 19 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 20 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 21 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 22 },
        { "low_power_mode" : 1, "no_prefetch" : true, "affine_to_cpu" : 23 },
   ],

The reason I am not using CPU #0 core 1 is because I have that core affined to the Nvidia GPU's.

My overall hash rate for the CPU's only has jumped from 827 to 947. Power only increased 32 watts so it is profitable to use those extra 120 H/s.
https://www.cryptocompare.com/mining/calculator/xmr?HashingPower=120&HashingUnit=H%2Fs&PowerConsumption=32&CostPerkWh=0.074&MiningPoolFee=1

My system now produces 1640 H/s while using 486 watts.

Profit per year now has increased to $510.
https://www.cryptocompare.com/mining/calculator/xmr?HashingPower=1640&HashingUnit=H%2Fs&PowerConsumption=486&CostPerkWh=0.074&MiningPoolFee=1

Hey. I'll tell you my secret. I sold all my mining equipment long ago, even though I didn't have to pay for electricity. Mining has long been a profitable business. Those who had at one time to build a huge farm, and they will continue to earn,Yu long since paid off the cost of the equipment.

This would correlate to the 6344's I tested.  A pair does 800h/s.  Xmr-stak default config is 9T per CPU with 3T at low power for double mem and consumption of all L2/L3.  In dual configuration this is drawing 250w from the wall.  I tried setting 24T and 1x on memory usage but that was 40H/s slower and same power draw.

Sundownz don't like me.  I just wanna be fwiends and swap knowledge. 

Not sure who this is directed to?

But for me the Supermicro Server with dual Opteron 6348's and three Nvidia GTX 750's produces 1545 H/s and uses 426 watts. My electric cost is 7.4 cents per KWh.

The Monero profitability chart show a yearly profit of $465.

https://www.cryptocompare.com/mining/calculator/xmr?HashingPower=1545&HashingUnit=H%2Fs&PowerConsumption=426&CostPerkWh=0.074&MiningPoolFee=1

The Supermicro Server only cost me $252 and the three GTX 750's $120 for a total of $372. Payback is about 10 months. Realize though this hardware has value even if not for mining and that if I were to resell the server I would get well over $600 for it alone.

is it worth??
 how about hashrate??

Been going slower than I had hoped... been very busy with my main business.

But the 5th rack of GPU rigs is 100% done and the 2nd rack of Dell servers is 50% done.

I am starting on the 6th rack of GPU rigs tomorrow.

Hi All.

I was following this thread since some time ago, why? I've started to mine using z400 workstations. The system have a strange behavior. I have a rig with 3x 1070 gtx, mining eth using claymore dual, I know, maybe there are more profitable options, but, eth algo on nvidia cards allows to reduce power consumption, heat and noise, and for me this is a key point also.

The strange behavior is that when I boot up the system and I start to mine using claymore miner, at the beggining, for some hours, between 12 and 24 hours the system seems to be faster than the "estimated hashrate" in +15 +20% this happened lot of times, after some hours the system performance goes down, and starts to makes peaks and valleys.

I've tested conservative oc for the memory, just +400 to discard this is a problem of the oc settings, I've tested in two different z400 workstations, with different risers, and the behavior is always the same. The gpu's are all the time under 60ºC, so I don`t think is a temp problem.

At this point I have no clue about where can be the problem. so any help would be appreciated.

I know there's something called luck, but I've seen this behavior lot of times, so is not somthing that happens radomly.

Thanks in advance.

I use ethOS, so I do all my setting adjustments in the config files. But 580s run hot anyway, I can't get mine down below around 115 or so...

boy i really gotta get around to learning how to mod my bios, my 4 rx580 rig is using like 700w 

Agreed. My 1060s run at around 75W each, and my 1070s run 95-105...

Thats not low power consumption lol at least not in 2018.

I've built a small miner myself too, without spending a whole fortune to do so. I'm using two AMD R7 260X (Overclocked from the factory) which I bought, used, for dirty cheap from Ebay (Something like $50-60 each). They don't feature high-end performance like the R9 Or Titan series but they have low power consumption (100-110 Watts on full load each). Therefore, having a high performance to power ratio.

I'm considering to purchase a few more of these in the future, depending on how well the GPU mining "industry" goes.