Topic: [ANN] Spondoolies-Tech - carrier grade, data center ready mining rigs - page 110. (Read 1260395 times)

Agree.
The logic of scaling is very complicated and takes into account 4-5 different parameters for optimisation. Also every ASIC is different and has different top freq` even at same voltage.

If that one goes hotter, the one behind it will drop completely out of the loop, eventually taking the whole loop down. The unit optimizes the freq for each chip so that it'll keep everything running. Don't be worried about one freq. Look at the whole unit.

yeah, you don't multiply .. you integrate (i.e. find the area under the consumption trend chart) over the desired time interval. There are numerical methods to do just that.

What I'd like to see added to the SP20's firmware is the ability to adjust the PSU efficiency multiplier. Using a Titanium rated PSU, I'm about 50-60w below the reported "wall watts" when measured with a kill-a-watt. No biggie, but it would be nice.

Does anyone know how to fix this on LOOP[0] ?
Even after restart of the miner it still on 760Hz-770Hz only.

Thanks. 

that is in the asic stats page (but cumulative kwh isnt)

Anyone know which power connectors on the back of the SP20s are front and rear of the cards?

Another nice feature would be to multiply the "wall watts" number shown in the ASIC stats page with time to get total kWh consumed..

Not really... Im only paying about .10 USD per kWh for electricity. About .065 if I can get a service change
Later this month. Besides, I like dealing with the machines.  Anyone have any luck w hosted miners?

19 KWh is high.  Have you thought about hosting?  I would guess you could pay around half of what your paying now.

I'm using 19 KWh per hour to generate 35 Ths. So, it looks like .5428 efficiency- best I can do with it. Can do better if I eliminate the AC on cold days (open bay doors).

Don't see that.

If you look below that, do you have a wall watts number.
On the SP20s it shows total watts at the wall, although it always shows about 5% high vs what they actually use measured by my killawatt.

This is what I'm showing on one of the ASIC stats page. showing 3543 Ghs. I'm relatively new to SP31 but seems highly efficient. I see 791w and 792w in brackets.  Is that the measured power output for Psus? Then I guess the 791 and 792 are averages.  This is the most efficient combination of settings I can find. If anyone can beat it let me know please.


Uptime:26250 | FPGA ver:62
-----BOARD-0-----
PSU[EMERSON1200]: 800->748w[748 748 748] (->791w[791 791 791]) (lim=750) 0c cooling:0/0x0
-----BOARD-1-----
PSU[EMERSON1200]: 784->748w[748 748 748] (->792w[792 792 792]) (lim=750) 0c cooling:0/0x0
LOOP[0] ON
 0: DC2DC/1/:[vlt1:589 vlt2:594(DCl:794 Tl:794 Ul:669) 42W  72A  57c] ASIC:[ 85c (120c) 635hz(BL: 635)  731 (E:193) F:0]
 1: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 42W  71A  57c] ASIC:[ 85c (120c) 630hz(BL: 630)  752 (E:193) F:0]
 2: DC2DC/1/:[vlt1:589 vlt2:594(DCl:794 Tl:794 Ul:669) 37W  63A  49c] ASIC:[ 85c (120c) 570hz(BL: 570)  654 (E:193) F:0]
LOOP[1] ON
 3: DC2DC/1/:[vlt1:593 vlt2:594(DCl:794 Tl:794 Ul:669) 36W  61A  55c] ASIC:[ 85c (120c) 545hz(BL: 545)  454 (E:192) F:0]
 4: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 41W  70A  64c] ASIC:[ 85c (120c) 620hz(BL: 620)  741 (E:193) F:0]
 5: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 43W  73A  65c] ASIC:[ 85c (120c) 655hz(BL: 655)  696 (E:193) F:0]
LOOP[2] ON
 6: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 32W  54A  60c] ASIC:[ 85c (120c) 485hz(BL: 485)  519 (E:188) F:0]
 7: DC2DC/1/:[vlt1:589 vlt2:594(DCl:794 Tl:794 Ul:669) 43W  74A  71c] ASIC:[ 85c (120c) 655hz(BL: 655)  766 (E:193) F:0]
 8: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 42W  72A  62c] ASIC:[ 85c (120c) 640hz(BL: 640)  744 (E:193) F:0]
LOOP[3] ON
 9: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 47W  79A  72c] ASIC:[ 85c (120c) 690hz(BL: 690)  798 (E:193) F:0]
10: DC2DC/1/:[vlt1:589 vlt2:592(DCl:794 Tl:794 Ul:669) 44W  74A  78c] ASIC:[ 85c (120c) 645hz(BL: 645)  766 (E:193) F:0]
11: DC2DC/1/:[vlt1:589 vlt2:592(DCl:794 Tl:794 Ul:669) 42W  72A  76c] ASIC:[ 85c (120c) 635hz(BL: 635)  730 (E:193) F:0]
LOOP[4] ON
12: DC2DC/1/:[vlt1:587 vlt2:592(DCl:794 Tl:794 Ul:669) 37W  63A  76c] ASIC:[ 85c (120c) 575hz(BL: 575)  687 (E:193) F:0]
13: DC2DC/1/:[vlt1:587 vlt2:592(DCl:794 Tl:794 Ul:669) 41W  69A  80c] ASIC:[ 85c (120c) 610hz(BL: 610)  738 (E:192) F:0]
14: DC2DC/1/:[vlt1:587 vlt2:592(DCl:794 Tl:794 Ul:669) 44W  75A  71c] ASIC:[ 85c (120c) 665hz(BL: 665)  797 (E:193) F:0]
LOOP[5] ON
15: DC2DC/1/:[vlt1:587 vlt2:592(DCl:794 Tl:794 Ul:669) 32W  55A  48c] ASIC:[ 85c (120c) 490hz(BL: 490)  613 (E:192) F:0]
16: DC2DC/1/:[vlt1:587 vlt2:592(DCl:794 Tl:794 Ul:669) 38W  65A  51c] ASIC:[ 85c (120c) 565hz(BL: 565)  615 (E:193) F:0]
17: DC2DC/1/:[vlt1:589 vlt2:592(DCl:794 Tl:794 Ul:669) 36W  62A  45c] ASIC:[ 85c (120c) 545hz(BL: 545)  585 (E:193) F:0]
LOOP[6] ON
18: DC2DC/1/:[vlt1:589 vlt2:594(DCl:794 Tl:794 Ul:669) 35W  60A  54c] ASIC:[ 85c (120c) 530hz(BL: 530)  592 (E:193) F:0]
19: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 43W  73A  64c] ASIC:[ 85c (120c) 635hz(BL: 635)  764 (E:193) F:0]
20: DC2DC/1/:[vlt1:589 vlt2:594(DCl:794 Tl:794 Ul:669) 40W  68A  62c] ASIC:[ 85c (120c) 590hz(BL: 590)  638 (E:193) F:0]
LOOP[7] ON
21: DC2DC/1/:[vlt1:589 vlt2:594(DCl:794 Tl:794 Ul:669) 44W  74A  69c] ASIC:[ 85c (120c) 640hz(BL: 640)  732 (E:193) F:0]
22: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 39W  65A  67c] ASIC:[ 85c (120c) 575hz(BL: 575)  647 (E:193) F:0]
23: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 44W  74A  60c] ASIC:[ 85c (120c) 640hz(BL: 640)  714 (E:193) F:0]
LOOP[8] ON
24: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 43W  72A  66c] ASIC:[ 85c (120c) 620hz(BL: 620)  706 (E:193) F:0]
25: DC2DC/1/:[vlt1:587 vlt2:592(DCl:794 Tl:794 Ul:669) 45W  76A  80c] ASIC:[ 85c (120c) 665hz(BL: 665)  740 (E:193) F:0]
26: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 43W  73A  77c] ASIC:[ 85c (120c) 650hz(BL: 650)  771 (E:193) F:0]
LOOP[9] ON
27: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 43W  73A  79c] ASIC:[ 85c (120c) 635hz(BL: 635)  752 (E:193) F:0]
28: DC2DC/1/:[vlt1:589 vlt2:594(DCl:794 Tl:794 Ul:669) 47W  80A  83c] ASIC:[ 85c (120c) 690hz(BL: 690)  841 (E:193) F:0]
29: DC2DC/1/:[vlt1:591 vlt2:594(DCl:794 Tl:794 Ul:669) 44W  74A  68c] ASIC:[ 85c (120c) 635hz(BL: 635)  709 (E:193) F

On the SP20's I can get way with 0.49x efficiency once I go below 1TH/s. I'm sure the SP3x can hit the same thing as it's the same guts just more of it. No?

Thats surprisingly good - i assume thats 1500W DC though, and closer to 1850W AC? generally the SP3X doesnt get more than 0.5W/GH efficient at the wall unless you really tune it low

Just window open a little, and the SP20s sucking air from there. Temps are around 60-70 at the loops.

I run my SP31s with the following settings, .64, .64, .67, and limit psus to 750w each.  I run fans at 40 with air conditioning and ceiling exhaust fans and get 3.5 Ths at about 1500w. Inlet temp 29c exhaust 69c. Running my entire operation at about 1/3 reduction, but savings on electrical makes up for it. Underclocking has made room for several more Sp-31s but can't find any!


uote author=miguelmorales85 link=topic=521520.msg10870134#msg10870134 date=1427199319]

a SP20 to '4' fan speed? wow. Is you office at freezing temperature? How much is the temp?
[/quote]

a SP20 to '4' fan speed? wow. Is you office at freezing temperature? How much is the temp?