[ANN] Spondoolies-Tech - carrier grade, data center ready mining rigs

[raskul]

he's winding you up. my guess is it'll not be 'til next summer, at least.

Perhaps for the generation after but the stats for the sp50 show 28nm 12th 6kw, so current gen with power savings. Hence assuming it shouldn't be too far away.

Where did you see such stats please?

[1715368944]

1715368944

[1715368944]

1715368944

[1715368944]

1715368944

[pak13]

he's winding you up. my guess is it'll not be 'til next summer, at least.

Perhaps for the generation after but the stats for the sp50 show 28nm 12th 6kw, so current gen with power savings. Hence assuming it shouldn't be too far away.

Where did you see such stats please?

One of a few places ... (speculation perhaps but the numbers look legit? We know they are working to get down to 0.2j/gh at 28nm, or lower)
MOD EDIT: Suspect link removed.

[raskul]

he's winding you up. my guess is it'll not be 'til next summer, at least.

Perhaps for the generation after but the stats for the sp50 show 28nm 12th 6kw, so current gen with power savings. Hence assuming it shouldn't be too far away.

Where did you see such stats please?

One of a few places ... (speculation perhaps but the numbers look legit? We know they are working to get down to 0.2j/gh at 28nm, or lower)
http://www.*scamshop_linkdeleted*.nl/product/sp50-yukon-bitcoin-miner/

oh deary me. not another one  

firstly, I hope that no-one has sent money to that NL website
secondly, that's an awfully bad photoshopping job

and look! they sell 'cloud mining' too, now that's a surprise isn't it?

[pak13]

Where did you see such stats please?

One of a few places ... (speculation perhaps but the numbers look legit? We know they are working to get down to 0.2j/gh at 28nm, or lower)
http://www.*scamshop_linkdeleted*.nl/product/sp50-yukon-bitcoin-miner/

oh deary me. not another one  

firstly, I hope that no-one has sent money to that NL website
secondly, that's an awfully bad photoshopping job

and look! they sell 'cloud mining' too, now that's a surprise isn't it?

Well with Hanukkah fast approaching, perhaps they could give us a little information for the new year. Never hurts to ask!

[DevonMiner]

Ah yes ... spotted in the wild .. the SP62 ... very rare, if not unheard of ...

[dogie]

Thanks for pointing it out and thanks to dogie for info and nice pictures.
For those NOT on a metric system using freedom units it comes down to ~13.717 lb and 9.25in X 9.45in X 19.5in for the full package and 5.12in X 5.83in X15.6in for the machine alone.

Fixed

[philipma1957]

SP50 announcement date?

sssssshhhhhhh 

LOL spk50 already, seems its a bad mistake to purchase mining rig afterall


philipma1957: let me know if you want to get extra units, have couple that i havent use, and problaby will never

send me a pm

[wh00per]

Hi Guys,

Can anyone help me out. 

When I run the "Devs" command on an SP30 miner (running most current firmware) I get the following output.
av=4445838.29,MHS
5s=5923594.28,MHS
1m=5187361.37,MHS
5m=4678246.84,MHS
15m=4526365.21

I notice that the av number seems to be the only number that is accurate.  Can anyone tell me what the other number are?  I would like to know why the 5s shows almost 6TH/s in this example.

Thanks guys.

Statistics 101?... That aside

As the measure time interval decreases variance increases e.g. A 5 second average reading has much higher chance to vary above and below the expected value than a 5 minute window.

This is in part due to the manner in which the mining software calculates hashrate.

A personal example. I have a particular miner which consistently performs at 500Gh/s both in cgminer avg and pool but I e observed the 5s avg vary from 200G to 1T

Ok So my real question is, does av show the current (live) hash rate? 

I use http://www.multiminerapp.com/ to monitor the SP30's and it generally works great.  The one problem is that it consistently reports the miners hash far to high (between 5 - 9 THs).

The software polls the miner via the CGMiner api with the DEVS command to get the stats.  I am trying to ascertain what in the returned data is the live hashing speed so that I can get the devs to ensure that is what is being recorded by there software.

I hope that makes sense.  Thanks again guys

The av value is the closest to the reality    It's a moving average, get it via CGminer's SUMMARY[] = "['MHS av']= ....... "

As an example .. compare here the values on the first chart: http://powerprice.info/SPT/graphic.php

CGMiner hashing speed is calculated via the number of accepted/rejected shares, the average share time etc. I won't go in details here, but it's mainly a function of "luck" .. layman's terms: if you find shares quick, you have a high hash-rate.

[wh00per]

One of a few places ... (speculation perhaps but the numbers look legit? We know they are working to get down to 0.2j/gh at 28nm, or lower)
http://www._________________.nl/product/sp50-yukon-bitcoin-miner/

I could give you some negative trust, just because you posted that link (and you still keep it there after a bunch of posts).. what do you think? Should I?

[Epoch]

Does anyone know what cables the SP35 ships with? The website description says it includes mounting ears, but nothing about the PSU plug types or adapter cables.

The Emerson-based SP30 came with 2 IEC C19-C14 power cords; wondering if the SP35 comes with something similar.

[dogie]

Does anyone know what cables the SP35 ships with? The website description says it includes mounting ears, but nothing about the PSU plug types or adapter cables.

The Emerson-based SP30 came with 2 IEC C19-C14 power cords; wondering if the SP35 comes with something similar.

I can let you know tomorrow if my unit clears customs in time to invoice me before loading time.

[Guy Corem]

he's winding you up. my guess is it'll not be 'til next summer, at least.

Perhaps for the generation after but the stats for the sp50 show 28nm 12th 6kw, so current gen with power savings. Hence assuming it shouldn't be too far away.

Where did you see such stats please?

One of a few places ...

The site URL you published is scam.

[Tupsu]

I could give you some negative trust, just because you posted that link (and you still keep it there after a bunch of posts).. what do you think? Should I?

Do not play with matches, you can burn your ovn "Trust house"   

[Biffa]

Does anyone know what cables the SP35 ships with? The website description says it includes mounting ears, but nothing about the PSU plug types or adapter cables.

The Emerson-based SP30 came with 2 IEC C19-C14 power cords; wondering if the SP35 comes with something similar.

I asked this last month and was told the SP31 and 35 come with the same power cables as the SP30 did. And my SP31 came with them.

[jtoomim]

Does anyone have any stats on how loud a SP31 is at high, medium, and low settings?

I'm considering getting a SP35, but SP10s were a bit too loud for my taste, so I'm hoping a SP31 / SP35 will be easier on the ears due to the larger fans.

I don't even mind running it at a lower speed if the sound is more bearable.

SP30/31/35s are not as loud as the SP10, but they're not quiet by any means. If you're using 120V, the power supply fans will have to run at higher RPMs which makes them louder. On 220V, they may be more bearable. Due to the small size and high speed of the SP10 fans, the noise the SP10 makes is relatively high-pitched and annoying; the SP30s' noise is lower in frequency and more bearable in tone.

[zvisha]

Does anyone know what cables the SP35 ships with? The website description says it includes mounting ears, but nothing about the PSU plug types or adapter cables.

The Emerson-based SP30 came with 2 IEC C19-C14 power cords; wondering if the SP35 comes with something similar.

SP30 come in 2 cable flavours depending on PSU type.
The SP35 has cables similar to SP10, I don't know if it ships with them.

[zvisha]

Some explanation about  ASIC stats (per clients request.)

[uptime of minergate and fpga version]
Uptime:6457 | FPGA ver:62
[this PSU not in power or not recognised]
-----BOARD-0-----
PSU[UNKNOWN]: 0->(11w/11w)[11 11 11] (->11w[11 11 11]) (lim=1275) 0c 0GH cooling:0/0x20000000
[this PSU intake 1346w, output 1274w. Last 3 measurements show 1272,1274 and 1258, board rate=2176GH]
-----BOARD-1-----
PSU[NEW-MURATA]: 1346->(1274w/1274w)[1272 1274 1258] (->1222w[1221 1221 1222]) (lim=1275) 0c 2176GH cooling:0/0x20000000
[Those loops (top) have no power. Most probably PSU disconnected]
LOOP[0] OFF TO:0 (test serial failed or something)
0: disabled (i2c BAD, btw!)
1: disabled (i2c BAD, btw!)
2: disabled (i2c BAD, btw!)
LOOP[1] OFF TO:0 (test serial failed or something)
3: disabled (i2c BAD, btw!)
4: disabled (i2c BAD, btw!)
5: disabled (i2c BAD, btw!)
LOOP[2] OFF TO:0 (test serial failed or something)
6: disabled (i2c BAD, btw!)
7: disabled (i2c BAD, btw!)
8: disabled (i2c BAD, btw!)
LOOP[3] OFF TO:0 (test serial failed or something)
9: disabled (i2c BAD, btw!)
10: disabled (i2c BAD, btw!)
11: disabled (i2c BAD, btw!)
LOOP[4] OFF TO:0 (test serial failed or something)
12: disabled (i2c BAD, btw!)
13: disabled (i2c BAD, btw!)
14: disabled (i2c BAD, btw!)
[ Those are bottom loops ]
LOOP[5] ON TO:0
[ This DC2DC configured to 682mv, but actually provides 679mv. The limit of the voltage is the minimum of DCl (DC2DC limit), Tl (temperature limit) and Ul (userset limit). The DC2DC limit and temperature limit are learned durring the runtime, but forgoten every 3 hours for case of environment temperature change. The ASIC pulls 68 watt, 92 amper and the DC2DC temperature is 59c. The ASIC is bellow 90C, while the top limit is 120c. The freq` is 710MHz, the limit FOR THIS VOLTAGE FOR THIS ASIC is 710MHz. it found 226 shares, and it has 193 functioning engines. The other flags are not important.]
15: DC2DC/1/:[vlt1:679 vlt2:682(DCl:794 Tl:794 Ul:749) 63W  92A  59c] ASIC:[ 85c (120c) 710hz(BL: 710)  226 (E:193) F:0 L:0]
16: DC2DC/1/:[vlt1:677 vlt2:682(DCl:794 Tl:794 Ul:749) 66W  97A  62c] ASIC:[ 85c (120c) 745hz(BL: 745)  220 (E:192) F:0 L:0]
17: DC2DC/1/:[vlt1:679 vlt2:682(DCl:794 Tl:794 Ul:749) 59W  87A  54c] ASIC:[ 85c (120c) 675hz(BL: 675)  191 (E:193) F:0 L:0]
LOOP[6] ON TO:0
18: DC2DC/1/:[vlt1:679 vlt2:682(DCl:794 Tl:794 Ul:749) 66W  97A  62c] ASIC:[ 85c (120c) 735hz(BL: 735)  246 (E:192) F:0 L:0]
19: DC2DC/1/:[vlt1:679 vlt2:682(DCl:794 Tl:794 Ul:749) 62W  91A  66c] ASIC:[ 85c (120c) 715hz(BL: 715)  232 (E:193) F:0 L:0]
20: DC2DC/1/:[vlt1:677 vlt2:682(DCl:794 Tl:794 Ul:749) 66W  97A  71c] ASIC:[ 85c (120c) 760hz(BL: 760)  240 (E:193) F:0 L:0]
LOOP[7] ON TO:0
21: DC2DC/1/:[vlt1:679 vlt2:682(DCl:794 Tl:794 Ul:749) 64W  95A  74c] ASIC:[ 85c (120c) 735hz(BL: 735)  225 (E:193) F:0 L:0]
22: DC2DC/1/:[vlt1:679 vlt2:685(DCl:794 Tl:794 Ul:749) 67W  99A  78c] ASIC:[105c (120c) 765hz(BL: 765)  227 (E:193) F:0 L:0]
23: DC2DC/1/:[vlt1:677 vlt2:685(DCl:794 Tl:794 Ul:749) 67W  99A  65c] ASIC:[105c (120c) 740hz(BL: 740)  235 (E:193) F:0 L:0]
LOOP[8] ON TO:0
24: DC2DC/1/:[vlt1:677 vlt2:682(DCl:794 Tl:794 Ul:749) 71W 105A  72c] ASIC:[110c (120c) 775hz(BL: 775)  210 (E:193) F:0 L:0]
25: DC2DC/1/:[vlt1:681 vlt2:682(DCl:794 Tl:794 Ul:749) 66W  96A  81c] ASIC:[110c (120c) 770hz(BL: 770)  240 (E:193) F:0 L:0]
26: DC2DC/1/:[vlt1:677 vlt2:682(DCl:794 Tl:794 Ul:749) 69W 101A  83c] ASIC:[105c (120c) 795hz(BL: 795)  225 (E:193) F:0 L:0]
LOOP[9] ON TO:0
27: DC2DC/1/:[vlt1:677 vlt2:682(DCl:794 Tl:794 Ul:749) 70W 102A  85c] ASIC:[110c (120c) 795hz(BL: 795)  246 (E:193) F:0 L:0]
28: DC2DC/1/:[vlt1:677 vlt2:682(DCl:794 Tl:794 Ul:749) 66W  97A  85c] ASIC:[110c (120c) 775hz(BL: 775)  236 (E:191) F:0 L:0]
29: DC2DC/1/:[vlt1:677 vlt2:680(DCl:794 Tl:680 Ul:749) 70W 104A  72c] ASIC:[115c (120c) 785hz(BL: 785)  233 (E:193) F:0 L:0]

[Rate, wattage summery of DC2DCs, A:asic count, TMP: input output temperature]
[H:HW:2176Gh (0),W:999,L:0,A:15,MMtmp:0 TMP:(31)=>=>=>(0,76)]
[state of queues, not important]
Pushed 27 jobs , in HW queue 4 jobs (sw:2, hw:2)!
[statistics, wins:total shares, last minute shares, hw errors (wrong hash)]
min:49 wins:3350[this/last min:21/27] bist-fail:410, hw-err:1
[Leading zeroes, how much idle cycles over last secon/minute, computed rate, shares per board per minute]
leading-zeroes:42 idle promils[s/m]:0/0, rate:1887gh/s asic-count:1620 (wins:0+21)
Fan:80, conseq:200
AC2DC BAD: 0 0
[Running time / Idle time]
R/NR: 6376/0
RTF asics: 0
[Fet types (not important for you) and error counters (nothing to be afraid of, errors handled internally and do not effect rate)]
FET: 0:255 1:0
 0 restarted      0 reset          0 reset2         0 fake_wins
 0 stuck_bist     0 low_power      0 stuck_pll      0 runtime_dsble
 0 purge_queue    0 read_timeouts  0 dc2dc_i2c       0 read_tmout2    0 read_crptn
 0 purge_queue3   0 bad_idle
 1 err_murata
Adapter queues: rsp=0, req=22

[edonkey]

Thanks for the ASIC stats info. That's helpful. Maybe it can make it's way into the help text?

[Guy Corem]

https://www.youtube.com/watch?v=z4FUk6TXK18

[edonkey]

I've got free power in a data center, so I'm interested in getting the most out of my SP20. Noise and power use are not a concern.

With the stock settings I get about 1.675 GHs out of the unit. Unfortunately I'm not having much luck overclocking.

What I did was take my stock max voltage of 0.75 and increased by 0.01 then waited an hour for things to settle down. At 0.76 I got a very slight increase from 1.675 to 1.685, but any attempt to increase beyond that resulted in less than the original hash rate.

I didn't notice any obvious differences in the ASIC temperatures. Although since the current GUI just shows a snapshot of the current temperature, it's kind of difficult to get a sense of temperature trends. That is, if the blade is ramping up then hits the max temp, it will back down, but you can't really see that without historical temperature info.

Anyway, after searching around some more it sounds like maybe my unit was already hitting its maximum temperature on some blades and backing down. So I reverted my max voltage to the original 0.75, then turned up the fan to 90. Just increasing the fan resulted in about the same hash rate increase as bumping up to 0.76.

I think what's going on is that some of my blades are hitting the max temp and backing down, with the original manufacturer's settings. In fact I can see in the hash rate graph that it spikes up near 1.7 GHs then backs down periodically. It may well be that I simply cannot push it any further without additional cooling.

Here's a snapshot of my current ASIC stats:

Code:

Uptime:25033 | FPGA ver:100
-----BOARD-0-----
PSU[UNKNOWN]: 0->286w[286 286 286] (->286w[286 286 286]) (lim=288) 0c cooling:0/0xcc
-----BOARD-1-----
PSU[UNKNOWN]: 0->272w[272 272 272] (->272w[272 272 272]) (lim=288) 0c cooling:0/0xcc
-----BOARD-2-----
PSU[UNKNOWN]: 0->283w[283 283 283] (->283w[283 283 283]) (lim=288) 0c cooling:0/0xcc
-----BOARD-3-----
PSU[UNKNOWN]: 0->262w[262 262 262] (->262w[262 262 262]) (lim=288) 0c cooling:0/0xcc
LOOP[0] ON TO:0
 0: DC2DC/1/:[vlt1:732 vlt2:738(DCl:794 Tl:794 Ul:749) 116W 158A  64c] ASIC:[110c (125c) 1115hz(BL:1115) 1236 (E:193) F:0 L:0]
 1: DC2DC/1/:[vlt1:732 vlt2:741(DCl:794 Tl:794 Ul:749) 115W 155A  82c] ASIC:[ 85c (125c) 1115hz(BL:1115) 1171 (E:193) F:0 L:0]
LOOP[1] ON TO:0
 2: DC2DC/1/:[vlt1:732 vlt2:735(DCl:794 Tl:735 Ul:749) 116W 158A 100c] ASIC:[120c (125c) 1115hz(BL:1115) 1170 (E:193) F:0 L:0]
 3: DC2DC/1/:[vlt1:705 vlt2:709(DCl:794 Tl:709 Ul:749) 104W 147A 100c] ASIC:[110c (125c) 1065hz(BL:1065) 1186 (E:193) F:0 L:0]
LOOP[2] ON TO:0
 4: DC2DC/1/:[vlt1:728 vlt2:733(DCl:794 Tl:794 Ul:749) 112W 152A  65c] ASIC:[ 85c (125c) 1090hz(BL:1090) 1188 (E:193) F:0 L:0]
 5: DC2DC/1/:[vlt1:726 vlt2:733(DCl:794 Tl:794 Ul:749) 117W 160A  82c] ASIC:[100c (125c) 1095hz(BL:1095) 1259 (E:193) F:0 L:0]
LOOP[3] ON TO:0
 6: DC2DC/1/:[vlt1:718 vlt2:722(DCl:722 Tl:725 Ul:749) 112W 155A  91c] ASIC:[120c (125c) 1085hz(BL:1085) 1143 (E:193) F:0 L:0]
 7: DC2DC/1/:[vlt1:679 vlt2:688(DCl:794 Tl:688 Ul:749) 100W 146A  94c] ASIC:[120c (125c) 1040hz(BL:1040) 1165 (E:193) F:0 L:0]

[H:HW:1682Gh,W:895,L:0,A:8,MMtmp:0 TMP:(21)=>=>=>(75,71)]
Pushed 29 jobs , in HW queue 4 jobs (sw:2, hw:2)!
min:40 wins:9518[this/last min:13/21] bist-fail:448, hw-err:0
leading-zeroes:42 idle promils[s/m]:0/0, rate:1468gh/s asic-count:3320 (wins:10+3)
Fan:90, conseq:200
AC2DC BAD: 0 0
R/NR: 24820/0
RTF asics: 0

Note that three are at 120c. They appear as yellow in the web GUI. My guess is that these are the ones that are bumping up to the 125c limit and backing down, but without historical data I can't be sure. Maybe there's a log event when an ASIC backs off? If so, I don't know what to look for in the event log.

I guess that I can try setting the fan to 100. Noise in the data center is not a problem. But with cold air blowing directly on the front of the unit (measured as 20 °C on the SP20 Temp Front sensor), it's somewhat disconcerting that I'm seeing the ASICs overheating at stock settings.

Any suggestions on how I can maximize the hash rate from this unit would be appreciated.

Also, I should note that I've seen at least one case where cgminer restarted, apparently on its own. It happened last night around 1 AM, according to the dashboard's CGMiner Uptime value. The main unit itself did not reboot, just cgminer.

I looked through the event log but could not find an event corresponding with the cgminer restart. Although it's kind of hard to deal with the event log because it's flooded with "DC2DC ASIC WARNING" entries for blades 2 and 3.

What causes the cgminer to reboot? Is that normal?