Swedish ASIC miner company kncminer.com

[murraypaul]

the only way to win is to bet on all horses

If all the options produce less Bitcoins than they cost to buy and run, then the only way to win is ... not to play.

That's only if you follow the crowd.

Well the crowd here is buying ASICs, so following them does seem like a bad idea
Edit: Paying for would be a better phrase. Buying somehow implies that you actually get something.

[Mota]

AVALON seems to be releasing GEN2 (add that to hashfast, ActiveMining, Bitmine.ch, Cointerra etc...) do you think my % increase in my calculations was optimistic?

where? certainly not on their website.

[naRky]

AVALON seems to be releasing GEN2 (add that to hashfast, ActiveMining, Bitmine.ch, Cointerra etc...) do you think my % increase in my calculations was optimistic?

where? certainly not on their website.

http://avalon-asics.com

[plasmoske]

AVALON seems to be releasing GEN2 (add that to hashfast, ActiveMining, Bitmine.ch, Cointerra etc...) do you think my % increase in my calculations was optimistic?

where? certainly not on their website.

Tradehill auction.

[r3animation]

http://www.eligius.st/~wizkid057/newstats/userstats.php/17Czc8RVL3FU5T2MLx2zLbRnpfBNgH9vFo

1.2 TH/s. Ouch.

[btc_uzr]

3 hours   705.63 Gh/s   1774352
22.5 minutes   693.91 Gh/s   218112
256 seconds   1,324.77 Gh/s   78963
128 seconds   1,456.37 Gh/s   43403

seems like a second Jupiter is online now
and first one is doing great

[kano]

April 08 to Sep 30 is 4 months? Right. It's 5.75 months ... anyone got their September delivery yet?

If it wasn't cgminer, I'd wonder about the hash rate being correct ...
... except ... the driver can get it wrong easily enough.
They should have displayed 2^32 * A / Elapsed after about an hour ... hopefully they do run for an hour ...
... unless they don't know how cgminer works ... oh right they only wrote a 1 or 2k line driver, they probably didn't know about the rest of the other 40k+ lines of code in there ...

I doubt they would provide official numbers on a product they're going to release which would then be heavily scrutinized against said numbers, without verifying their data so they don't end up with a foot in their mouth.

No idea, but we've seen a lot of this happen ...
I guess we'll find out when someone finally gets one and reports the numbers.

OK, so just to point out that they really don't understand what the numbers there on the screen mean ...
It's hashing at 469GH/s

Picture: (click to see bigger)


Firstly the screen says: 497.0GH/s (the number on the right is what the driver produces)

Calculation:
Start = 00:06:40
Share that just happened says: 00:28:46
Elapsed = 22:06 = 1326 seconds

Accepted + Rejected = 144587 + 260 = 144847 1diff shares

Hash rate = 2^32 * 1diff / Elapsed
= 2^32 * 144847 / 1326
= 469,165,254,844 or 469GH/s

So yeah they really should learn what the numbers mean ... ... ... ...
... and I have assumed the Rejects are valid shares and included them in the hash rate ... though that's a guess, they may not be valid?

Also note:
HW: 28517 is an awesome ... 28517 / (28517 + 144587 + 260) = 28517/173364 = 16.46% ...

Edit: lastly that WU: number there at the top of 8001.3 means 8001.3 * 2^32 / 60  = 572,755,363,758
So they are losing >100GH/s of work vs the replies they are getting back from the device ...

[arlekyn13]

So anyone can estimate the daily production of units at 22/30 stations?

[timmmers]

So anyone can estimate the daily production of units at 22/30 stations?

10 a day per station would seem very very low and would be 300 a day. They said all October orders would be out mid Oct. and I'd easily believe that.
Our roving reporter will know better no doubt

[ur0pl]

And what about the 5V connector.

Should I buy a 230V to 5V DC converter?

yeah there's a lot of think that has to be sorted out imho, let me summarize my own list:

- Is a 850W PSU really enough for a Jupiter? (in the 'orama video (1) I saw the power consumption at wall was 843W)
- Do we need a 5V connector?
- What is the delivery ETA for day 1? Or looking at it in another way how long does it take to do the last tweaking to the machine?

(1) http://www.youtube.com/watch?feature=player_detailpage&v=CiQRxJuoRR8#t=25

You can later buy a second psu to power units if need be. 850 w /. 85 efficiency psu = can handle a wall draw of 1k watts.

[vesperwillow]

So anyone can estimate the daily production of units at 22/30 stations?

10 a day per station would seem very very low and would be 300 a day. They said all October orders would be out mid Oct. and I'd easily believe that.
Our roving reporter will know better no doubt

He already said it takes less than 2 minutes to build a Jupiter, which is what a lot of us presumed to begin with--it's just plugging together modular items. Board, board, cable, cable, etc.

[kano]

So anywhere up to 300 a day you think?

At 469GH/s ( less than a BFL Minirig ) that's 140TH a day

Current difficulty is 148,819,199.80509263 = 1065.289327 TH/s

So if they ONLY spend a week making them, difficulty will go up ... 92%

So the next 2 difficulty changes should together be over 184% ...

Lets see who doesn't stop mining in 4 weeks

So lets say they only make them for 2 weeks - 184% - what will the return on them be? ...

[The Avenger]

http://eligius.st/~wizkid057/newstats/userstats.php/17Czc8RVL3FU5T2MLx2zLbRnpfBNgH9vFo

Woo! First entire bitcoin mined on a Jupiter!

This is my 3rd and final dibs call on the first 1BTC mined on a jupiter!! (1st call here https://bitcointalk.org/index.php?topic=170332.msg3273425#msg3273425 & 2nd call https://bitcointalk.org/index.php?topic=170332.msg3274904#msg3274904)

My BTC address below, thanks KNC      
\/
\/
\/

(I will report back if KNC play by the sacred and unbreakable rules of calling dibs on something with humourous homage to this moment of mining history )

[Xialla]

So lets say they only make them for 2 weeks - 184% - what will the return on them be? ...

hey Kano, here is very strong base of blind KNC fans, really recommend you to stop this kind of discussion.)

[mmitech]

So anywhere up to 300 a day you think?

At 469GH/s ( less than a BFL Minirig ) that's 140TH a day

Current difficulty is 148,819,199.80509263 = 1065.289327 TH/s

So if they ONLY spend a week making them, difficulty will go up ... 92%

So the next 2 difficulty changes should together be over 184% ...

Lets see who doesn't stop mining in 4 weeks

So lets say they only make them for 2 weeks - 184% - what will the return on them be? ...

I was thinking about the same thing, BTW do not forget that bitfurystrikesback is massively shipping October pre-orders, some 400GH/s devices are coming online as well.
add to them what BFL and ASICMINER is shipping and you get some interesting times coming soon  

[arlekyn13]

So I guess my probably last order for Oct delivery from Sep 1st will be delivered next week? After the difficulty increase (ouch!) though...

[Mota]

So anywhere up to 300 a day you think?

At 469GH/s ( less than a BFL Minirig ) that's 140TH a day

Current difficulty is 148,819,199.80509263 = 1065.289327 TH/s

So if they ONLY spend a week making them, difficulty will go up ... 92%

So the next 2 difficulty changes should together be over 184% ...

Lets see who doesn't stop mining in 4 weeks

So lets say they only make them for 2 weeks - 184% - what will the return on them be? ...

I was thinking about the same thing, BTW do not forget that bitfurystrikesback is massively shipping October pre-orders, some 400GH/s devices are coming online as well.
add to them what BFL and ASICMINER is selling and you get some interesting times coming soon  

Still not convinced that BFL will ship anything worthwile and in numbers....

[vesperwillow]

So lets say they only make them for 2 weeks - 184% - what will the return on them be? ...

hey Kano, here is very strong base of blindness KNC fans, really recommend you to stop this kind of discussion.)

Lol, it truthfully sucks more people don't look into the numbers this way. I've maintained for awhile that difficulty is going to get stupid. For some reason, folks believe KNC only had a few orders. The magnitude of their actual order quantity is going to send them into shock. Kano's numbers are extremely conservative compared to reality.

The most profitable mining is occuring at this very moment. It's all downhill once the UPS trucks arrive at folks' doorstep around the world.

BFL isn't a major player anylonger, even with the few large miners they have to ship. The Monarch isn't a reality for awhile, aside from the lack of tapeout success, they haven't worked out the heating issues and are already projecting higher power consumption (lolz).

[Its About Sharing]

April 08 to Sep 30 is 4 months? Right. It's 5.75 months ... anyone got their September delivery yet?

If it wasn't cgminer, I'd wonder about the hash rate being correct ...
... except ... the driver can get it wrong easily enough.
They should have displayed 2^32 * A / Elapsed after about an hour ... hopefully they do run for an hour ...
... unless they don't know how cgminer works ... oh right they only wrote a 1 or 2k line driver, they probably didn't know about the rest of the other 40k+ lines of code in there ...

I doubt they would provide official numbers on a product they're going to release which would then be heavily scrutinized against said numbers, without verifying their data so they don't end up with a foot in their mouth.

No idea, but we've seen a lot of this happen ...
I guess we'll find out when someone finally gets one and reports the numbers.

OK, so just to point out that they really don't understand what the numbers there on the screen mean ...
It's hashing at 469GH/s

Picture: (click to see bigger)


Firstly the screen says: 497.0GH/s (the number on the right is what the driver produces)

Calculation:
Start = 00:06:40
Share that just happened says: 00:28:46
Elapsed = 22:06 = 1326 seconds

Accepted + Rejected = 144587 + 260 = 144847 1diff shares

Hash rate = 2^32 * 1diff / Elapsed
= 2^32 * 144847 / 1326
= 469,165,254,844 or 469GH/s

So yeah they really should learn what the numbers mean ... ... ... ...
... and I have assumed the Rejects are valid shares and included them in the hash rate ... though that's a guess, they may not be valid?

Also note:
HW: 28517 is an awesome ... 28517 / (28517 + 144587 + 260) = 28517/173364 = 16.46% ...

Edit: lastly that WU: number there at the top of 8001.3 means 8001.3 * 2^32 / 60  = 572,755,363,758
So they are losing >100GH/s of work vs the replies they are getting back from the device ...

Thanks for sharing that.
I've said this phrase too often, but "Excuse my ignorance", shouldn't CGMiner display the actual mining rate without having to do calculations?
When I look at my erupters mining, it is displaying 333MHash/s each, plain as day. Why not here?

Thanks Kano,
IAS

[erk]

So anywhere up to 300 a day you think?

At 469GH/s ( less than a BFL Minirig ) that's 140TH a day

Current difficulty is 148,819,199.80509263 = 1065.289327 TH/s

So if they ONLY spend a week making them, difficulty will go up ... 92%

So the next 2 difficulty changes should together be over 184% ...

Lets see who doesn't stop mining in 4 weeks

So lets say they only make them for 2 weeks - 184% - what will the return on them be? ...

The 110nm miners and the BFL boys will be the first to switch off, well before KNCminer users. I wouldn't want to be holding ASICminer shares right now.