Swedish ASIC miner company kncminer.com

[bitpop]

Well you forgot one thing, knc has no track record yet unlike the others. they have to sell at a loss to build trust first.

[bitpop]

And building a 1 ghs or 1 ths costs the same to them.

[Rampion]

I wanted to check out if kncminer is legit... so i made these thoughts:

The Jupiter has 350GH/S and sells for $7000. Vat is included already and swedish vat is 25%. 25% of $7000 means $5250 is the real selling price because the rest doesnt belong to kncminer, it belongs to the state.

Asicminer claims that they can bring 1TH/s online for costs of $10k. The knc Jupiter has 350GH/s which would mean AM would pay $3500 for the same hashpower getting online. KNCMiner sells the same for $5250. Its not only a blade like with AM, its a complete miner. That means kncminer makes $1750 win only when getting $7000. And that is correct only if they can create a complete miner as cheap as Asicminer a bare erupter blade creates. And AM buys in huge quantities.

So they really sell the miner for so low surcharge? Where do they produce the asics and assemble them? Can it be as cheap as in china where AM produces?

Avalon is taking 101,24BTC for 85GH. That would mean 416BTC for 350GH/s. Or $52100. Avalon... kncminer takes only $7000($5250). Thats 10%.

Asicminer is selling 10GH/s Erupter blades for 50BTC. 1750BTC would that mean for 350GH/s. Or $218750. But kncminer throws its miner away for 2.4% of that?

Then... the avalon chips 10K are selling for 782,1BTC. Each has 282MH/s. Thats 2820GH/s for 10K chips. 350GH/s would mean 1242 Avalon Asic Chips. That would mean 97,14BTC price with avalon for chips having 350GH/s. Or $12142.5. KNCMiner takes for the same hashpower only less than the half price. And avalon is only selling the chips for that price, no miner included...

Im not sure what i should think about it... can kncminer really create miners for prices similar to Asicminer? And they practically throw it away for nothing while resellers could make a fortune with these miners? I tend to think thats a bit strange... i mean are they even more bitcoin friendly than avalon? And avalon is giving away their hashpower for low already... resellers make a fortune with their miners.

For me this doesnt add up at the moment. Any thoughts?

If it's true they are doing 28nm its possible. I'm pretty much sure that they are legit and in good faith (ORSoC is a 9 years old, small but established company) - but you could say the same things of BFL. And the delivery dates KNC is promising are VERY ambitious: September for the first 500 and October for everyone else. They say they can do huge volume once they start up the engines

This is so good that its very difficult to believe, if they achieve that they will be golden, as they will sell miners likes there's no tomorrow. Still, we are all quite burnt out by the BFL situation, and that's why we are trying to run all the checks possible, including personal visits to their facilities, company checks, etc.

If someone is thinking about doing very big orders, I would recommend exploring the possibility of having a contract with them, to at least consider some penalties if they deliver with a delay bigger of let's say 30 days. If they can comply the requirements needed to run as a legitimate business and not as an endless pre-order scheme that offers no guarantees whatsoever (like BFL), I'm sure ORSoC's 2012 turnover will go x100.

[papamoi]

just my two cents


competition are good for the consumer prices so i m happy with knc announced price

[Biomech]

I wanted to check out if kncminer is legit... so i made these thoughts:

The Jupiter has 350GH/S and sells for $7000. Vat is included already and swedish vat is 25%. 25% of $7000 means $5250 is the real selling price because the rest doesnt belong to kncminer, it belongs to will be stolen by the state.

FTFY

As to the costs involved, I'm not in the business but I've been around a lot of guys that do produce chips. The only really large cost is the mask, as I understand it, and that's pretty much a one time deal. Also, given the size of ASIC chips, it is likely that the manufacturer could include them on a mask for other things, bringing the overall cost down quite a bit as it saves waste on the wafer.

Now, I could be completely wrong about that last bit, it's old news (like a decade old) but I know it used to be done that way for small run things. Merge it with another product run.

Of course, they could have just raised sufficient capital to build a full mask and hope to recoup it through sales and services. Unless and until they ship and/or detail their business model to us, the latter of which they have no obligation to do, we're all just shooting in the dark.

[bitpop]

Xoxox knc

[Bitcoinorama]

From speaking to Sam, pricing was determined by non-recurring engineering and their long term goal. That's their priority to prove they have the capability of developing 28nm chips, manufacture en mass and crucially cover NRE in the initial run of Jupiter and Saturn.

After which point the products will sell themselves. They can introduce a variety of miners at a variety of price points all using the same chips. This means that the network can be distributed thoroughly and people of a variety of financial standing can become involved. Once NRE is covered, the cost to acquire large volumes and produce becomes significantly less expensive, meaning they can acquire a profit and offer a range of products and earn from multiple revenue streams.

The chips themselves are not expensive the NRE is where the expense is. Economies of scale driven by interest in a proven technology and no further reoccurring engineering costs will make mining far more attractive price wise then where it is today. Avalon's bulk chip pricing does not reflect what they are paying for chips at all. An ASIC costs between $0.50 and $2 to manufacture per chip. Avalon has stated a desire to cover NRE for their next gen chip, but thy are making a very healthy profit until they have some competition. Avalon are just determining the retail price point because they got there first and people are prepared to pay it. An FPGA is expensive as it's a fully customisable chip via software. FPGAs cost between a few dollars and a few hundred dollars each. An ASIC is a predefined chip for the most part. The initial layers are a standard design with only the final few masks customisable within the wafer.

FPGAs are great for runs under 10k, Asics Over 10k. Finalise the design in FPGA and then cement it at produce in volume at a lot less cost as an ASIC. Jupiter/Saturn will still feature an FPGA as a programable chip on the board that controls the subsequent ASICs.

Ignore the pricing you see on any current genuine ASIC devices they are a mix of passing NRE onto consumers and just profiting as much as they can based on demand. ASICminer and Avalon especially.

[dan99]

I believe Avalon are making tons of money now since they already recover their ROI and NRE way back ... got good profit and now on their next gen chip. With Knc in the picture is going to give good competition and benefits the users..

[SebastianJu]

First two avalon batches were priced in dollar and both with a price of 1500$ circa.

Still higher than kncminer... for $5250 buying these old avalons would mean getting 220,5GH/s.

Well you forgot one thing, knc has no track record yet unlike the others. they have to sell at a loss to build trust first.

You would do this when you start a business? I mean you only need to show a running miner and you dont have to sell with loss anymore... or give out samples to Kano or similar.

And building a 1 ghs or 1 ths costs the same to them.

Why do you think this? The chips has to be created... and when Asicminer now, as the masks are all done still have costs of $10000 per 1TH online... and KNCMiner is at the start only?

@Rampion

So the smaller die size leads to this... if thats true they could outcompete everyone...

Running checks in bitcoin world is soooo important... i had to learn this the hard way too...

@Biomech... i wonder... Friedcat claimed in the past that it would be better for them not to create a better asic but instead build more of the olds only. The reason friedcat stated was the development cost of course and the higher production price of smaller die chips. I dont know really how big this additional production cost would be. I mean when a chip gets the double hashrate because of smaller die but the production cost per chip is doubled too it wont make sense... it wont be that bad of course... but there must be a reasonable effect so that friedcat decided to go with that old technique instead...

@dan99 But kncminer cant exclude the NRE either... they have to get it back one way or the other to be successful. So i dont really see that kncminer could sell the miners without getting back nre from the start... there is no low sell needed. They only would need to show a working miner or give samples to kano and co and the miners would sell from itself.

I think time will tell... the visit that comes now only will show working fpgas as far as i see... so seeing a working asic chip will be a good thing for sure.

When i see the plans for septembre and octobre... they do this the first time... every asic-dev had problems first to build the actual miner... bfl has the worst problems... so i wonder... Asicminer and Avalon claimed they will develop a better chip now too. Im interested to see how this race will end.

[Bitcoinorama]

I believe Avalon are making tons of money now since they already recover their ROI and NRE way back ... got good profit and now on their next gen chip. With Knc in the picture is going to give good competition and benefits the users..

This is the problem however;

This is very interesting and the guy in the video absolutely nails it at the end.  It's a vlog about BFL's mistakes and the 'white space' within the Bitcoin industry.

The whitespace is particularly referencing bringing professionalism to the current amateur (albeit some very intelligent engineers) run companies that aren't operating as businesses and are struggling with issues of reputation, security, and delivering on time.

It's not a dig at BFL et al. per se, but highlights that the problems stem from a general lack of business acumen based on prior experience with dealing with the public, and states the importance of degree of professionalism required to exist in a global marketplace. It's where Avalon fail badly as well. That said, from the videos of Yifu at Bitcoin 2013 last week, it's where Avalon have no intention of improving. As their guys have no interest in any administrative or customer relations, purely development. I can't blame them, they are engineers and development is more interesting, where as admin and cs are boring and restrictive, but seeing as you need customers to continue development it cannot be overlooked and it undoubtedly is the current void in mining manufacture.

It concludes that it will continue to be an issue, but less so over time as this is a burgeoning sector.

http://youtu.be/8vwuwibQ_tQ

In essence there is a massive opportunity here to buck the trend and lead with a strong example. KNC please take note!

[Bitcoinorama]

First two avalon batches were priced in dollar and both with a price of 1500$ circa.

Still higher than kncminer... for $5250 buying these old avalons would mean getting 220,5GH/s.

Well you forgot one thing, knc has no track record yet unlike the others. they have to sell at a loss to build trust first.

You would do this when you start a business? I mean you only need to show a running miner and you dont have to sell with loss anymore... or give out samples to Kano or similar.

And building a 1 ghs or 1 ths costs the same to them.

Why do you think this? The chips has to be created... and when Asicminer now, as the masks are all done still have costs of $10000 per 1TH online... and KNCMiner is at the start only?

@Rampion

So the smaller die size leads to this... if thats true they could outcompete everyone...

Running checks in bitcoin world is soooo important... i had to learn this the hard way too...

@Biomech... i wonder... Friedcat claimed in the past that it would be better for them not to create a better asic but instead build more of the olds only. The reason friedcat stated was the development cost of course and the higher production price of smaller die chips. I dont know really how big this additional production cost would be. I mean when a chip gets the double hashrate because of smaller die but the production cost per chip is doubled too it wont make sense... it wont be that bad of course... but there must be a reasonable effect so that friedcat decided to go with that old technique instead...

@dan99 But kncminer cant exclude the NRE either... they have to get it back one way or the other to be successful. So i dont really see that kncminer could sell the miners without getting back nre from the start... there is no low sell needed. They only would need to show a working miner or give samples to kano and co and the miners would sell from itself.

I think time will tell... the visit that comes now only will show working fpgas as far as i see... so seeing a working asic chip will be a good thing for sure.

When i see the plans for septembre and octobre... they do this the first time... every asic-dev had problems first to build the actual miner... bfl has the worst problems... so i wonder... Asicminer and Avalon claimed they will develop a better chip now too. Im interested to see how this race will end.

ASICs are expensive to develop initially, cheap to manufacture and limited once the custom design is finalised as to how they function. FPGAs are fully customisable at any point, they are expensive to manufacture, but offer the broadest functional ability which is why you use them as a starting point.

Avalon are making around 500% profit on every chip sold. They aren't saints, they are in this for the cash. Their product pricing is relative to what they feel they can squeeze from demand with Bitcoin's uncertain future. They cover the NRE for their next gen and fund their lifestyles. They're not spending it on admin and customer services.

ASIC miner are selling old stock at vastly greater than cost to satisfy an immediate demand with an already manufactured product. They then use the profit to better their own gains and develop next gen for themselves. They will then sell that off once they've used it for themselves at a price to those that must have it now. They don't do pre-orders as they built the tech for themselves and for their own use. They then sell it second hand to the highest bidder.

BFL are not designers or engineers. They priced fair initially, but then made mistake after mistake at great cost. I still cannot believe they mass manufactured casing (albeit the same casing used for their FPGAs) to fulfil their pre-orders and had it on standby without having working prototypes to know whether the design would realistically fit and remain stable at the power/temp demands required/initially suggested.

@Sebastian KNC's NRE is covered with this initial production, they already stated that. It's not like they are starting from scratch and have to forge new business contacts with ORSoC on board. It's the entire reason they've partnered, they are using the OpenCores route to development.

Here's some additional communication I had a couple of weeks back (I have X'd out the manufacturer as I'm not sure whether I'm allowed to mention them), the "--" is a break between a couple of emails;


As for the price drop you assumed. What I meant to say is that we don’t make any money unless we can sell around 500 devices, We only start to get profit after that number of devise has been sold.
 
The ORSoC and Kennemar & Cole relationship has been talked about by many and it’s not a secret but people have made assumptions so I will clear it up for you
 
Me and Andreas (Kennemar & Cole)
Started the idea, contacted ORSoC and showed them our plans, told them we would only do this if they joined us on the venture, a simple customer/supplier relationship was not strong enough for this purpose.
 
So we all started (all 5 of us) a brand new company called KNCMiner. So KncMiner consists of 5 directors. Front of house is us and back of house is ORSoC (best of both worlds) essentially ORSoC don’t get anything unless the product is successful and neither do me or Andreas, so we are all in it 100% and that’s the only way it would of worked.
--
And I agree with you about the Litecoin miner and it is definitely our most likely road. We have Jupiter and Saturn which take care of the upper end of the market for miners, as of today we don’t have a product for people who have a few hundred quid to spend, we know that. We brought the entry down with Saturn and once we have enough cash to cover the huge NRE XXXXXXXX are making us pay. We may very well offer up a product that is a few hundred quid allowing even more people to take part. Then we can truly service customers who today can buy a graphics card and compete. There will be nothing for them to buy in the summer months that will makes sense and I don’t think that fair. We have to do all we can to stop this becoming a rich man’s game.


So take what you will from that...

[dan99]

First two avalon batches were priced in dollar and both with a price of 1500$ circa.

Still higher than kncminer... for $5250 buying these old avalons would mean getting 220,5GH/s.

Well you forgot one thing, knc has no track record yet unlike the others. they have to sell at a loss to build trust first.

You would do this when you start a business? I mean you only need to show a running miner and you dont have to sell with loss anymore... or give out samples to Kano or similar.

And building a 1 ghs or 1 ths costs the same to them.

Why do you think this? The chips has to be created... and when Asicminer now, as the masks are all done still have costs of $10000 per 1TH online... and KNCMiner is at the start only?

@Rampion

So the smaller die size leads to this... if thats true they could outcompete everyone...

Running checks in bitcoin world is soooo important... i had to learn this the hard way too...

@Biomech... i wonder... Friedcat claimed in the past that it would be better for them not to create a better asic but instead build more of the olds only. The reason friedcat stated was the development cost of course and the higher production price of smaller die chips. I dont know really how big this additional production cost would be. I mean when a chip gets the double hashrate because of smaller die but the production cost per chip is doubled too it wont make sense... it wont be that bad of course... but there must be a reasonable effect so that friedcat decided to go with that old technique instead...

@dan99 But kncminer cant exclude the NRE either... they have to get it back one way or the other to be successful. So i dont really see that kncminer could sell the miners without getting back nre from the start... there is no low sell needed. They only would need to show a working miner or give samples to kano and co and the miners would sell from itself.

I think time will tell... the visit that comes now only will show working fpgas as far as i see... so seeing a working asic chip will be a good thing for sure.

When i see the plans for septembre and octobre... they do this the first time... every asic-dev had problems first to build the actual miner... bfl has the worst problems... so i wonder... Asicminer and Avalon claimed they will develop a better chip now too. Im interested to see how this race will end.

ASICs are expensive to develop initially, cheap to manufacture and limited once the custom design is finalised as to how they function. FPGAs are fully customisable at any point, they are expensive to manufacture, but offer the broadest functional ability which is why you use them as a starting point.

Avalon are making around 500% profit on every chip sold. They aren't saints, they are in this for the cash. Their product pricing is relative to what they feel they can squeeze from demand with Bitcoin's uncertain future. They cover the NRE for their next gen and fund their lifestyles. They're not spending it on admin and customer services.

ASIC miner are selling old stock at vastly greater than cost to satisfy an immediate demand with an already manufactured product. They then use the profit to better their own gains and develop next gen for themselves. They will then sell that off once they've used it for themselves at a price to those that must have it now. They don't do pre-orders as they built the tech for themselves and for their own use. They then sell it second hand to the highest bidder.

BFL are not designers or engineers. They priced fair initially, but then made mistake after mistake at great cost. I still cannot believe they mass manufactured casing (albeit the same casing used for their FPGAs) to fulfil their pre-orders and had it on standby without having working prototypes to know whether the design would realistically fit and remain stable at the power/temp demands required/initially suggested.

@Sebastian KNC's NRE is covered with this initial production, they already stated that. It's not like they are starting from scratch and have to forge new business contacts with ORSoC on board. It's the entire reason they've partnered, they are using the OpenCores route to development.

Here's some additional communication I had a couple of weeks back (I have X'd out the manufacturer as I'm not sure whether I'm allowed to mention them), the "--" is a break between a couple of emails;


As for the price drop you assumed. What I meant to say is that we don’t make any money unless we can sell around 500 devices, We only start to get profit after that number of devise has been sold.
 
The ORSoC and Kennemar & Cole relationship has been talked about by many and it’s not a secret but people have made assumptions so I will clear it up for you
 
Me and Andreas (Kennemar & Cole)
Started the idea, contacted ORSoC and showed them our plans, told them we would only do this if they joined us on the venture, a simple customer/supplier relationship was not strong enough for this purpose.
 
So we all started (all 5 of us) a brand new company called KNCMiner. So KncMiner consists of 5 directors. Front of house is us and back of house is ORSoC (best of both worlds) essentially ORSoC don’t get anything unless the product is successful and neither do me or Andreas, so we are all in it 100% and that’s the only way it would of worked.
--
And I agree with you about the Litecoin miner and it is definitely our most likely road. We have Jupiter and Saturn which take care of the upper end of the market for miners, as of today we don’t have a product for people who have a few hundred quid to spend, we know that. We brought the entry down with Saturn and once we have enough cash to cover the huge NRE XXXXXXXX are making us pay. We may very well offer up a product that is a few hundred quid allowing even more people to take part. Then we can truly service customers who today can buy a graphics card and compete. There will be nothing for them to buy in the summer months that will makes sense and I don’t think that fair. We have to do all we can to stop this becoming a rich man’s game.


So take what you will from that...

good sharing ...

[Phinnaeus Gage]

I'm surprised there's not SSL cert. on their login page.

Do they need one if Paypal handles payments and card info?

Nah! Guess not! I'll just continue to log into their unprotected site.

[Phinnaeus Gage]

We may very well offer up a product that is a few hundred quid allowing even more people to take part.

Why did they use quid instead of dollars or euros, being that they're located in the Euro-Zone with the US probably being one of their biggest markets?

[canth]

Europeans tend to travel and be more aware of foreign currencies than us Americans. It's slang anyway - could refer to Euros for all it matters.

[Bitcoinorama]

We may very well offer up a product that is a few hundred quid allowing even more people to take part.

Why did they use quid instead of dollars or euros, being that they're located in the Euro-Zone with the US probably being one of their biggest markets?

They are selling in dollars.

Sam is English, as am I.

Quid is slang for a British pound, which is approximately equal to a Euro anyway. I'd use it off the top of my head to refer to a pound coin or note. It's like you saying bucks, I don't think the euro has a slang term, not a polite one anyway!



Ironically what people don't realise is the meaning of the term 'pound' and where it came from. It refers to imperial weights, and it meant the bearer of the note (a contract, a promissory note, an IOU) held the equivalent value of a pound in weight of sterling silver (92.5% Silver, 7.5% Nickel) in a vault in her Bank of England. There are 2.2 pounds in a kilo, it's like 454g I think.

Anyway the most dishonest part of it is on every note is the queen's face printed alongside the following quote on every denomination of note stating; 'I promise to pay the bearer the sum of..." £10, £20 etc. Obviously she'll never honour that promise, nor will the BoE.

(Scottish note, but still British sterling - at least until next year!)


That's fraud and hyper inflation for you!

Viva la Bitcoin!!

[Phinnaeus Gage]

The quid question was answered satisfactory.

[dan99]

Still no news on the date of the opening day, one more day left for the week...

[Miracek]

with the changes in the initial offerings of products, how does one find out where exactly they are in the list of "pre-orders"?  before someone yells that there are no pre-orders since no money has been exchanged, thanks I am aware of that, I am just using the terminology used on the website.

[Twh]

with the changes in the initial offerings of products, how does one find out where exactly they are in the list of "pre-orders"?  before someone yells that there are no pre-orders since no money has been exchanged, thanks I am aware of that, I am just using the terminology used on the website.

Well you did get an order ID, which approximately tells you how many people have pre-ordered before you. The number is very seriously inflated since you don't need to pay anything, I'd go as far as saying that it could easily be 5-15 times more than what these people pre-ordering could actually buy. The order ID is also generic, so it includes all orders, even Mars units - basically all 3 miners they have on the website.
The issue is that if they actually start delivering, all the people who registered 20 pre-orders just because they could at the beginning are possibly going to try to transfer these - sell them. If KnCMiner somehow manages to prevent that, then that's great, but I don't see how they could do it without just starting from scratch on all orders all over again.