Spondooliestech SP20


#1

Ok, this miner has a nominal hashrate of 1.7Th/s +/- 10%

I’ve been trying to reconcile the power consumption figures, as I see low values but they conflict when you work them out.

Headline power consumption = 0.34W/GHs typical

PSU needed 1200W

Now this needs examining further. 1700GHs x 0.34 = 578W

Huh ?

Aah, there are 8 asics and the headline power consumption is per asic

But each asic is 200GHs

So 200 x 0.34 = 68W per asic x 8 = 544W

Huh ?!

I can’t work out what voltage they are using for the calculations.

Aah, according to this review it pulls 1310W at 110v, so on 220v it should only need 655W.
That’s closer to the calculated rate but not exact.

I’m left bewildered and confused about a suitable PSU for this miner.

Help ? :dizzy_face:


#2

I think they over-spec the SP20 a little bit ( Marketing team ? )

You can refer to Scott review for power consumption as good value.

Power consumption stay the same on 110V or 220V.
At 220V you have a little efficiency gain but that’s all.

Power = Voltage * Current

Going 220V you draw haft the current compare to 110V.
Main advantage is lower current, reducing risk of overheat in wiring…


#3

Thanks :smile:
Doesn’t explain why the power per GHs is out though.
1700GHs x 0.34 = 578W whatever voltage is used.

Going backwards, 1200W / 1700 GHs = 0.706 W/GHs

The S5 uses 590W at the wall according to Bitmain.

590W / 1150 GHs = 0.513 W/GHs which is what Bitmain state.

So where do Spondoolies get 0.34 W/GHs from ?

It is touted as the most efficient miner, but it can’t be if it requires more W per GH than the S5.

That makes it slightly more efficient than the S3+ which currently isn’t profitable here in the UK. The S3+ costs more to run then it makes. (edited this as the S3+ works out to 0.789 W/GHs)

2 x SP20 = 2400W = 3400GHs (hardware costs GBP = 2 x £400 = £800 + PSUs)
3 x S5 = 1800W = 3450GHs (hardware costs GBP = 3 x £320 = £960 + PSUs)

I can get a whole extra TH of S5 for the same energy as 2 x SP20s :open_mouth:

I guess the savings are made in the hardware costs, but that’s not a huge difference in the medium to long term. (get the S5 hardware costs back in saved energy in ~3 months.)

[edit - added hardware costs]
Trying to understand why I should get an SP20 over an S5 :slight_smile:


#4

Am curious too, Miner performances does not meet that claim.

ATM S5 is the best miner at 0.5W/GH

SP20 is better build, more finish product.
Emissions Compliance FCC / CE Safety Certification CE
Cost more

S5 is open box product to save production cost
It can’t be certified FCC / CE. EMI is not contain and is ESD sensitive
China doesn’t care much about regulation.

But for power vs cost you can’t beat them lol


#5

:smile:
I just got thinking because people are advertising the 0.34 figure on ebay and it just didn’t add up. Guess that’s what happens when you copy & paste your hardware specs :wink:


#6

One of the things too that messes up power consumption figures is the PSU. PSUs have some loss and depending on their quality you can get some variance as well from what your miner is actually using and what the PSU is pulling.
I recommend getting one of these for getting real numbers on your consumption.
Watt Meter NewEgg
This is one of the main meters I use.
It should help you out.


#7

I have one of these, an EMS 2.0 In-Home Energy Monitor. Pretty good, and that’s how I can trust the Bitmain figures. The peak draw “at the wall” (S5) has been 610W but mostly its around 590 to 595W on an AX1200i PSU.


#8

Nice.
I am using EVGA 1300 SuperNova’s as well as for the SP20in my review a customer server PSU with custom breakout board they sent me to test out.