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

[philipma1957]

I am driving my Kia Forte at 105 miles per hour and it seems like a rough ride can anyone help me.

Yeah don't push  the car drive it at 85 miles per hour.

Except in this case Phillip, it would be more alike to purchasing a Mustang that is manufacturer rated for 20 MPG @ 80 MPH, but won't go over 60 MPH.  If the dealer told them "don't drive over 60", there would be some upset customers.

I have wrote a hundred times the sp20 is not really a 1700 miner.  If not 100 times more then 25 times. maybe 50 or 60 times.

It will do 1500 with cold air and high fans.

and is best in the 1100 to 1400 range.

If he wants the chip to stay under  120c he has to run it slower.  or completely tear the unit down and add big heatsinks with more cooling.

And spondoolies wrote 1700 gh -+ 10%  at .61watts a gh  .

 so  1700 - 10% = 1530gh  and he gets 1563gh.  and he gets .63 watts a gh maybe .64 watts so he is within spec.

 he burns tons of power to do it. and if his power is not 5 cents or less he runs his gear in a wasteful way.  (mining btc) mining other coins I do not know.

This is old news.  I do not understand why miners max the gear and complain it overheats.  over clock is a losing method results in dead gear.

[dmeter]

 so  1700 - 10% = 1530gh  and he gets 1563gh.  and he gets .63 watts a gh maybe .64 watts so he is within spec.

 he burns tons of power to do it. and if his power is not 5 cents or less he runs his gear in a wasteful way.  (mining btc) mining other coins I do not know.

This is old news.  I do not understand why miners max the gear and complain it overheats.  over clock is a losing method results in dead gear.

I write my power is not 5 cent i pay eletric -13$ cent if i dont burn 5-7 kw by hour i pay penality 0,15 $ if i burn 5 kw i receive 0,13 $ kwh to other account.
do you hear negative electric price.this is EU Intelligent Electrical Power Grids. In summer when come turist i must slow down speed.
next year if i dont bay new bitcoin miner i plug electric  heat and heat atmosfere in street,or change led light whit 12x 500W lamp.

Negative prices are a price signal on the power wholesale market that occurs when a high inflexible power generation meets low demand. Inflexible power sources can’t be shut down and restarted in a quick and cost-efficient manner. Renewables do count in, as they are dependent from external factors (wind, sun).
Negative prices are not a theoretical concept. Buyers are actually getting money and electricity from sellers. However, you need to keep in mind that if a producer is willing to accept negative prices, this means it is less expensive for him to keep their power plants online than to shut them down and restart them later.

Also you " over clock is a losing method results in dead gear"
If i put my miner in bed and cover whit silk blanked and cut off power plug any 50-28 nm miner is dead for 3-6-9 month

[padrino]

Its still hard for me the belive that sp20s made SP-tech no money.

-The chips are already developed and ready
-Boards are hillariously cheap .  It has 3 boards basically if you count the controller.
-No power supply.

The only thing I can think is the cost comes from the steel casing , if thats the case strip it down and sell the miner for the same price with just the hashing boards and a fan.  And charge extra
for an encased board.

The only oher thing i can think is they don't want competion for thier new bulk hardware which is strange to me since the hardware is not even in production yet?

Have you ever manufactured semi conductors? Please enlighten me on how boards are "hilariously cheap", pull one apart and price up the components to start to get yourself a feel for it...

[kendog77]

Its still hard for me the belive that sp20s made SP-tech no money.

-The chips are already developed and ready
-Boards are hillariously cheap .  It has 3 boards basically if you count the controller.
-No power supply.

The only thing I can think is the cost comes from the steel casing , if thats the case strip it down and sell the miner for the same price with just the hashing boards and a fan.  And charge extra
for an encased board.

The only oher thing i can think is they don't want competion for thier new bulk hardware which is strange to me since the hardware is not even in production yet?

Have you ever manufactured semi conductors? Please enlighten me on how boards are "hilariously cheap", pull one apart and price up the components to start to get yourself a feel for it...

Guy seems like an honest, stand up guy, so when he says SPTech didn't make money on the SP20 I believe him.

[jmumich]

Its still hard for me the belive that sp20s made SP-tech no money.

-The chips are already developed and ready
-Boards are hillariously cheap .  It has 3 boards basically if you count the controller.
-No power supply.

The only thing I can think is the cost comes from the steel casing , if thats the case strip it down and sell the miner for the same price with just the hashing boards and a fan.  And charge extra
for an encased board.

The only oher thing i can think is they don't want competion for thier new bulk hardware which is strange to me since the hardware is not even in production yet?

Have you ever manufactured semi conductors? Please enlighten me on how boards are "hilariously cheap", pull one apart and price up the components to start to get yourself a feel for it...

Guy seems like an honest, stand up guy, so when he says SPTech didn't make money on the SP20 I believe him.

I agree, though regardless, if SPTech made money on them, then it would stand to reason that they'd make a similar model using next-gen chips. Unless they don't like money.

[djroadking]

So my SP10 just arrived (ordered from Ebay) and so far it's made me decide not to buy anything from the website again. The case is banged up to hell to the point that the power supply wouldn't even slide out. The unit was not well protected even in the cardboard box... So does anyone have any tips for opening up an sp10? I'm going to take the casing and see if I can bend the dents out.

[thomas_s]

So my SP10 just arrived (ordered from Ebay) and so far it's made me decide not to buy anything from the website again. The case is banged up to hell to the point that the power supply wouldn't even slide out. The unit was not well protected even in the cardboard box... So does anyone have any tips for opening up an sp10? I'm going to take the casing and see if I can bend the dents out.

The boxes aren't really meant for double/multiple shipping times. After the first shipment the box usually gets banged up a little but not much, but it weakens the packaging enough for the next shipment to just fall apart. If it came by pallet originally you get 1 extra shipment out of it usually. I wouldn't buy a SP10 in original packaging I would want it / make sure its wrapped in bubble wrap and potentially a new box.

[Marvell1]

Anyone seen a flashing green light on an SP30 psu ?  on the UI see this

Code:

Uptime:133 | FPGA ver:62 | BIST in 12
-----BOARD-0-----
PSU[EMERSON1200]: 1024->(960w/960w)[960 960 960] (->1042w[1030 1036 1042]) (lim=1275) 0c 1831GH cooling:0/0x0
-----BOARD-1-----
PSU[EMERSON1200]: 64->(0w/0w)[0 0 0] (->11w[11 11 11]) (lim=1275) 0c 0GH cooling:0/0x0
LOOP[0] ON TO:0 (w:16)
 0: DC2DC/1/:[vlt1:662 vlt2:669(DCl:794 Tl:794 Ul:749) 55W  80A  41c] ASIC:[ 85c (125c) 600hz(BL: 600)    5 (E:193) F:0 L:0]
 1: DC2DC/1/:[vlt1:664 vlt2:669(DCl:794 Tl:794 Ul:749) 55W  79A  43c] ASIC:[ 85c (125c) 600hz(BL: 600)    7 (E:193) F:0 L:0]
 2: DC2DC/1/:[vlt1:660 vlt2:669(DCl:794 Tl:794 Ul:749) 61W  88A  41c] ASIC:[ 85c (125c) 660hz(BL: 660)    4 (E:193) F:0 L:0]
LOOP[1] ON TO:0 (w:11)
 3: DC2DC/1/:[vlt1:666 vlt2:669(DCl:794 Tl:794 Ul:749) 59W  85A  45c] ASIC:[ 85c (125c) 630hz(BL: 630)    2 (E:193) F:0 L:0]
 4: DC2DC/1/:[vlt1:660 vlt2:669(DCl:794 Tl:794 Ul:749) 56W  81A  48c] ASIC:[ 85c (125c) 610hz(BL: 610)    4 (E:193) F:0 L:0]
 5: DC2DC/1/:[vlt1:664 vlt2:669(DCl:794 Tl:794 Ul:749) 58W  84A  49c] ASIC:[ 85c (125c) 630hz(BL: 630)    5 (E:193) F:0 L:0]
LOOP[2] ON TO:0 (w:19)
 6: DC2DC/1/:[vlt1:664 vlt2:669(DCl:794 Tl:794 Ul:749) 52W  78A  53c] ASIC:[ 85c (125c) 620hz(BL: 620)    7 (E:193) F:0 L:0]
 7: DC2DC/1/:[vlt1:664 vlt2:669(DCl:794 Tl:794 Ul:749) 53W  79A  54c] ASIC:[ 85c (125c) 630hz(BL: 630)    6 (E:193) F:0 L:0]
 8: DC2DC/1/:[vlt1:666 vlt2:669(DCl:794 Tl:794 Ul:749) 53W  80A  48c] ASIC:[ 85c (125c) 620hz(BL: 620)    6 (E:193) F:0 L:0]
LOOP[3] ON TO:0 (w:11)
 9: DC2DC/1/:[vlt1:664 vlt2:669(DCl:794 Tl:794 Ul:749) 53W  80A  53c] ASIC:[ 85c (125c) 640hz(BL: 640)    3 (E:191) F:0 L:0]
10: DC2DC/1/:[vlt1:662 vlt2:669(DCl:794 Tl:794 Ul:749) 56W  84A  57c] ASIC:[ 85c (125c) 660hz(BL: 660)    4 (E:193) F:0 L:0]
11: DC2DC/1/:[vlt1:666 vlt2:669(DCl:794 Tl:794 Ul:749) 59W  89A  56c] ASIC:[ 85c (125c) 690hz(BL: 690)    4 (E:193) F:0 L:0]
LOOP[4] ON TO:0 (w:9)
12: DC2DC/1/:[vlt1:666 vlt2:669(DCl:794 Tl:794 Ul:749) 49W  74A  58c] ASIC:[ 85c (125c) 600hz(BL: 600)    5 (E:193) F:0 L:0]
13: DC2DC/1/:[vlt1:664 vlt2:669(DCl:794 Tl:794 Ul:749) 54W  81A  60c] ASIC:[ 85c (125c) 650hz(BL: 650)    3 (E:193) F:0 L:0]
14: DC2DC/1/:[vlt1:662 vlt2:669(DCl:794 Tl:794 Ul:749) 60W  87A  51c] ASIC:[ 85c (125c) 650hz(BL: 650)    1 (E:193) F:0 L:0]
LOOP[5] OFF TO:0 (test serial failed)
15: disabled (no power to DC2DC)
16: disabled (no power to DC2DC)
17: disabled (no power to DC2DC)
LOOP[6] OFF TO:0 (test serial failed)
18: disabled (no power to DC2DC)
19: disabled (no power to DC2DC)
20: disabled (no power to DC2DC)
LOOP[7] OFF TO:0 (test serial failed)
21: disabled (no power to DC2DC)
22: disabled (no power to DC2DC)
23: disabled (no power to DC2DC)
LOOP[8] OFF TO:0 (test serial failed)
24: disabled (no power to DC2DC)
25: disabled (no power to DC2DC)
26: disabled (no power to DC2DC)
LOOP[9] OFF TO:0 (test serial failed)
27: disabled (no power to DC2DC)
28: disabled (no power to DC2DC)
29: disabled (no power to DC2DC)

[H:HW:1831Gh (3000),DC-W:841,L:0,A:15,MMtmp:0 TMP:(21/21)=>=>=>(47/47 , 0/0)]
Pushed 32 jobs , in HW queue 4 jobs (sw:1, hw:1)!
min:54 wins:66[this/last min:33/33] bist-fail:23, hw-err:0
leading-zeroes:41 idle promils[s/m]:0/3, rate:1153gh/s asic-count:45 (wins:66+0)
wall watts:1088

[Marvell1]

Its still hard for me the belive that sp20s made SP-tech no money.

-The chips are already developed and ready
-Boards are hillariously cheap .  It has 3 boards basically if you count the controller.
-No power supply.

The only thing I can think is the cost comes from the steel casing , if thats the case strip it down and sell the miner for the same price with just the hashing boards and a fan.  And charge extra
for an encased board.

The only oher thing i can think is they don't want competion for thier new bulk hardware which is strange to me since the hardware is not even in production yet?

Have you ever manufactured semi conductors? Please enlighten me on how boards are "hilariously cheap", pull one apart and price up the components to start to get yourself a feel for it...

Guy seems like an honest, stand up guy, so when he says SPTech didn't make money on the SP20 I believe him.

I agree, though regardless, if SPTech made money on them, then it would stand to reason that they'd make a similar model using next-gen chips. Unless they don't like money.

Their saying the definetly WILL NOT produce consumer models for their next gen.  Thing is right now Bitman is sold out on their s4s SP20E is sold out
where the F are we supposed to buy miners from now lol. 

I don't expect any new hardware from SP-TECH till around end of summer in any case and it wont be consumer grade either way

[Guy Corem]

...
Their saying the definetly WILL NOT produce consumer models for their next gen...

What I said was that our initial form factor for the 3rd gen will be suitable for bulk buyers.
Please read my last messages.

[klondike_bar]

Anyone seen a flashing green light on an SP30 psu ?  on the UI see this

Code:

-----BOARD-0-----
PSU[EMERSON1200]: 1024->(960w/960w)[960 960 960] (->1042w[1030 1036 1042]) (lim=1275) 0c 1831GH cooling:0/0x0
-----BOARD-1-----
PSU[EMERSON1200]: 64->(0w/0w)[0 0 0] (->11w[11 11 11]) (lim=1275) 0c 0GH cooling:0/0x0

problem is the PSU - try unplugging it and letting it cool down for 30min. if that doesnt solve it the PSU might be dead. most likely your 1275W limit was a bit too high, and tripped the psu protection - this is fixed by changing the limit to ~1020W and letting it have the 30min cooldown period before starting it up again

[dmeter]

What I said was that our initial form factor for the 3rd gen will be suitable for bulk buyers.
Please read my last messages.

Please simple answer do your company want sale me 3rd gen miner SP-100 16 kw like 3 pcs nano farm SP-100    250-300 th or mini farm 15x SP100
whitout power supply and whitout Rackmount Cases.

[Marvell1]

Anyone seen a flashing green light on an SP30 psu ?  on the UI see this

Code:

-----BOARD-0-----
PSU[EMERSON1200]: 1024->(960w/960w)[960 960 960] (->1042w[1030 1036 1042]) (lim=1275) 0c 1831GH cooling:0/0x0
-----BOARD-1-----
PSU[EMERSON1200]: 64->(0w/0w)[0 0 0] (->11w[11 11 11]) (lim=1275) 0c 0GH cooling:0/0x0

problem is the PSU - try unplugging it and letting it cool down for 30min. if that doesnt solve it the PSU might be dead. most likely your 1275W limit was a bit too high, and tripped the psu protection - this is fixed by changing the limit to ~1020W and letting it have the 30min cooldown period before starting it up again

ok ill give it a shot thanks

[Finksy]

I have wrote a hundred times the sp20 is not really a 1700 miner.  If not 100 times more then 25 times. maybe 50 or 60 times.

With respect Phillip, I'm aware of this and you are right, but you are not the manufacturer and that's my issue.  I just don't find it comforting that we are supposed to settle for less.  Generally speaking, +/- ratings would signify that the rated number falls in the middle.  How many 1,870 GH/s SP20's have we seen? For that matter, how many 1,700 GH/s SP20's have we seen?  As someone with cheap power, I was a little disappointed that unlike every other miner I have purchased, my 17x SP20's do not average out to the 1.7 TH/s, as the manufacturer claimed.  I hate to drag other manufacturers into this, but when the other "big guys" claim a stated hashrate with their retail miners, that's not over-clocked and strung to the max, that is out of the box performance for months on end, and generally very consistent.  It is a different level of standard, that's all I'm saying.

I'm not unhappy with SP-Tech, I'm just disappointed with how they rated the SP20.

[philipma1957]

I have wrote a hundred times the sp20 is not really a 1700 miner.  If not 100 times more then 25 times. maybe 50 or 60 times.

With respect Phillip, I'm aware of this and you are right, but you are not the manufacturer and that's my issue.  I just don't find it comforting that we are supposed to settle for less.  Generally speaking, +/- ratings would signify that the rated number falls in the middle.  How many 1,870 GH/s SP20's have we seen? For that matter, how many 1,700 GH/s SP20's have we seen?  As someone with cheap power, I was a little disappointed that unlike every other miner I have purchased, my 17x SP20's do not average out to the 1.7 TH/s, as the manufacturer claimed.  I hate to drag other manufacturers into this, but when the other "big guys" claim a stated hashrate with their retail miners, that's not over-clocked and strung to the max, that is out of the box performance for months on end, and generally very consistent.  It is a different level of standard, that's all I'm saying.

I'm not unhappy with SP-Tech, I'm just disappointed with how they rated the SP20.

100% agreement with you on the bold type.  They should have rated it 1500 +-5% top speed  with addition under-clocking possible.

Then if you got above 1575 you would feel you did well.     Their    estimate   was Truly  a non-Scotty  estimate (obscure Star Trek the next generation reference).

[jmumich]

I have wrote a hundred times the sp20 is not really a 1700 miner.  If not 100 times more then 25 times. maybe 50 or 60 times.

With respect Phillip, I'm aware of this and you are right, but you are not the manufacturer and that's my issue.  I just don't find it comforting that we are supposed to settle for less.  Generally speaking, +/- ratings would signify that the rated number falls in the middle.  How many 1,870 GH/s SP20's have we seen? For that matter, how many 1,700 GH/s SP20's have we seen?  As someone with cheap power, I was a little disappointed that unlike every other miner I have purchased, my 17x SP20's do not average out to the 1.7 TH/s, as the manufacturer claimed.  I hate to drag other manufacturers into this, but when the other "big guys" claim a stated hashrate with their retail miners, that's not over-clocked and strung to the max, that is out of the box performance for months on end, and generally very consistent.  It is a different level of standard, that's all I'm saying.

I'm not unhappy with SP-Tech, I'm just disappointed with how they rated the SP20.

100% agreement with you on the bold type.  They should have rated it 1500 +-5% top speed  with addition under-clocking possible.

Then if you got above 1575 you would feel you did well.     Their    estimate   was Truly  a non-Scotty  estimate (obscure Star Trek the next generation reference).

That, or even a lower level, may have been good for marketing the SP20 as an out-of-the-box at-home miner. The SP20 is not suitable for use in most homes directly out of the box (way too loud), and it takes some time to calibrate them to get them just right depending on where they're placed, the ambient temperature, and I'm sure other factors. But it can be a good at-home miner. If it were marketed at say 1200, in most homes it would run with a fan speed of about 6-12% on auto, and that's tolerable. And it would have significant overclock potential (and some under).

Edit: Some recent GUI updates have made the SP20 a great home miner, in my opinion. If these features were developed further, and perhaps made more user friendly by, for example, allowing them to adjust to the proper environment perhaps (living room to 6% fan, garage to 60% fan, basement to 30% fan, etc), a next-gen unit directed to the home could sell really well. Edit2: This would make it easier for at-home miners to 'sell' to other household members: It's quiet, it's easy to use, it looks good, it might make a little money and if not it won't cost much. I suppose the SP20 would have to be able to be set to maximize its hashrate at a given fan speed.

[sloopy]

I haven't been able to make it back to this message. Crazy work days and nights
Good to talk to ya Phil, I hope things are going well.

My power isn't cheap, but I don't know what cheap is yet I'm so upside down in my mining, but it isn't a full-blown-standup-business either, it is more my sit-down-have-a-good-time-business-hobby.  
I like to know what each miner will do, and when one seemed different than the rest, I want to know why and if I can make it better.
I figured I would ask the pros you know? I do lean to the speed side from the years spent overclocking PCs with peltiers, hunks of copper, pumps, and I had a great time. It also feeds my technological appetite. I consume any type of input from real books, to any one of the PCs, istuff, droid this n that, although I rarely ever watch TV, listen to music mostly in my Kia. You really shouldn't make fun of my Kia Phil, words hurt  
...but I hear you, and have found the same challenge with underclocking, so I may be asking you why I can't do such and such on a certain system.

I don't know if it is a personality trait, or just a simple quirk (maybe annoying to people?)
I ask for opinions from people all the time. From things I sincerely know nothing about, to things I have been doing for 25 years, but I enjoy listening or reading how things are explained, how to do something your way, how would you apply the heatsink? Do you care if the Thermal material is conductive? Why? Why is my favorite. I do try to stay in the realm of topic so as to follow the rules and take my random questions PM.

I think there is a solution for every problem. Every single one. It is When enough resources are applied to a problem the solution is discovered. Many times the price of those resources is beyond what anyone will pay.

There are also things I have done over the years because I knew it was the correct thing to do, but do not fully understand it or have not experienced understanding yet. As I get older it is not acceptable in my personal life to do many things I don't understand, but there are some, back to that whole resources thing.

I have asked several people how they setup their home mining solutions in great detail. If they care to swap ideas, pictures, etc ... That whole two heads are better than one thing. I gravitate towards people who are willing to discuss any possibility, but also have a good idea of what the probabilities are. I have commented a couple of times about you Phil, I think you have a pretty sharp eye and good decision making skills and value your opinion about Bitcoin mining.

At the same time I have my own opinions, right or wrong, fresh n00b or not. I simply wanted to know what, if anything, others would do with a similar issue. I have not contacted Spondoolies because I have been slammed at work for the last 48 hrs and  I do not plan to keep that miner running at high hash unless, I decide I want to do so. In that scenario, I am looking for suggestions about the 120 chip.

I understand how much more power is used to squeeze out that last 50, 100 or 200 or 300 of hashpower on the SP20, S3, S3+, S4. The first thing I learned in this thread probably was underclocking is the name of this game now. It wasn't always though right?

If my line of questioning made someone think they should run their miner as hard as they can all the time I certainly did not intend to do so.

However, if anyone has done any heatsink mods for this specific scenario I would certainly enjoy chatting about it. We could also chat about the dream of one day having a need to run a miner as hard and fast as it can run. Similar to tuning a sports car, or in my case the lovely KiA  

I do prefer the Chevy though
  

I am driving my Kia Forte at 105 miles per hour and it seems like a rough ride can anyone help me.

Yeah don't push  the car drive it at 85 miles per hour.

These units have variance just like GPU's used to have. Try running this unit at 1400 gh not 1563 gh

use
0.680
0.680
0.680
0.680

0.690 max volts

240 watts
240
240
240

fan 80

BTC is like any other risky business. It also happens to be a fanatic passion among many people. I do not mean fanatic with a negative connotation, I am one myself ...but I hope a good one!

OnTopic:
I have a high temp and I'm not sure which chip(s) on which board to take a look at.

Loop 2 #4 I cannot seem to keep consistently cool, even with cold 0c air at the intake it stays higher at 120.
Please take a look and let me know your thoughts. I will contact Spondoolies warranty as well, but I've found discussing it here seems  fast and educational as well.

Ok enough of me blabberin
 
My AsicLoop NFO:

Code:

2.6.14
Uptime:29963 | FPGA ver:100 | BIST in 53
-----BOARD-0-----
PSU[UNKNOWN]: 0->(238w/238w)[238 238 238] (->238w[238 238 238]) (lim=288) 0c 382GH cooling:0/0xd0
-----BOARD-1-----
PSU[UNKNOWN]: 0->(247w/247w)[247 247 247] (->247w[247 247 247]) (lim=288) 0c 387GH cooling:0/0xd0
-----BOARD-2-----
PSU[UNKNOWN]: 0->(255w/255w)[255 255 255] (->255w[255 255 255]) (lim=288) 0c 399GH cooling:0/0xd0
-----BOARD-3-----
PSU[UNKNOWN]: 0->(245w/245w)[245 245 245] (->245w[245 245 245]) (lim=288) 0c 393GH cooling:0/0xd0
LOOP[0] ON TO:0 (w:2608)
 0: DC2DC/1/:[vlt1:707 vlt2:709(DCl:794 Tl:794 Ul:709) 97W 137A  59c] ASIC:[ 85c (125c) 990hz(BL: 990) 1308 (E:193) F:0 L:0]
 1: DC2DC/1/:[vlt1:703 vlt2:709(DCl:794 Tl:794 Ul:709) 95W 134A  77c] ASIC:[ 85c (125c) 990hz(BL: 990) 1300 (E:193) F:0 L:0]
LOOP[1] ON TO:0 (w:2696)
 2: DC2DC/1/:[vlt1:705 vlt2:709(DCl:794 Tl:794 Ul:709) 102W 143A  90c] ASIC:[105c (125c) 1020hz(BL:1020) 1398 (E:193) F:0 L:0]
 3: DC2DC/1/:[vlt1:707 vlt2:709(DCl:794 Tl:794 Ul:709) 97W 138A  92c] ASIC:[110c (125c) 990hz(BL: 990) 1298 (E:193) F:0 L:0]
LOOP[2] ON TO:0 (w:2762)
 4: DC2DC/1/:[vlt1:703 vlt2:709(DCl:794 Tl:794 Ul:709) 106W 150A  67c] ASIC:[120c (125c) 1060hz(BL:1060) 1383 (E:193) F:0 L:0]
 5: DC2DC/1/:[vlt1:703 vlt2:709(DCl:794 Tl:794 Ul:709) 99W 140A  81c] ASIC:[105c (125c) 1010hz(BL:1010) 1379 (E:193) F:0 L:0]
LOOP[3] ON TO:0 (w:2592)
 6: DC2DC/1/:[vlt1:701 vlt2:703(DCl:794 Tl:703 Ul:709) 100W 143A  89c] ASIC:[115c (125c) 1030hz(BL:1030) 1265 (E:193) F:1 L:0]
 7: DC2DC/1/:[vlt1:703 vlt2:709(DCl:794 Tl:794 Ul:709) 97W 137A  96c] ASIC:[115c (125c) 1010hz(BL:1010) 1327 (E:193) F:0 L:0]

[H:HW:1563Gh (500),DC-W:797,L:0,A:8,MMtmp:0 TMP:(25/25)=>=>=>(72/72 , 71/71)]
Pushed 10 jobs , in HW queue 4 jobs (sw:1, hw:1)!
min:18 wins:10658[this/last min:4/19] bist-fail:66, hw-err:0
leading-zeroes:42 idle promils[s/m]:0/0, rate:1328gh/s asic-count:4008 (wins:3+1)
wall watts:1081
Fan:90, conseq:200
AC2DC BAD: 0 0
R/NR: 29951/0
RTF asics: 0
FET: 0:5 1:5
 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
 0 err_murata
Adapter queues: rsp=1, req=24

Any and all facts, thoughts, opinions, and insults welcome 

[philipma1957]

@ sloopy

the chips do have variance.

I had a lot of GPU's running back in 2012.

20 to 24 of them.  Sometimes the particular chip in question is not good at over clock.

You can take  the gpu fully apart put good paste better heatsink and the gpu runs less then an identical gpu.


this heatsink  for gpus is really good.


http://www.newegg.com/Product/Product.aspx?Item=N82E16835242030&cm_re=gpu_heatsink-_-35-242-030-_-Product
It is so good that when I would mine with the gpus every one ran under 70c no matter what I did.

I have had quite a few gpus modded with these. and the exact same make and model gpus in the exact same rig can vary 10 %.

So your one hot chip may just be a variance  issue.  Now you can pull the heatsinks re paste them and maybe it is the answer.
  Or no matter what you do it will run hot at the setting you use.

[TheRealSteve]

They should have rated it 1500 +-5% top speed  with addition under-clocking possible.

Don't know about 1500, but 1600 seems like it would have been more reasonable at least.

Their    estimate   was Truly  a non-Scotty  estimate (obscure Star Trek the next generation reference).

"Oh, laddie.. you've got a lot to learn if you want people to think of you as a miracle worker!" - Scotty
Edit: And just for kicks, some research that counters that mantra

[Finksy]

"Your reputation is secure Scotty"