Topic: Swedish ASIC miner company kncminer.com - page 1805. (Read 3049501 times)

Please note that I've calculated the "worst case scenario", I can see even today that the network increase is slowing down

(I'm the author of the doc)

yep delivery date of miner is really critical and every day counts and cost BTC. according to this:

https://docs.google.com/spreadsheet/ccc?key=0AgSa2BKZz0WudHVlQVc0NmNVaXJEUHdLS090dWptSEE#gid=9

is difference between 15/9 and 30/9 more then 33BTC for jupiter.(

Completely. They never said they would deliver in the beginning of September, just September.

I didn't want to get sucked into Hashfast's words. I'm happy to see competition, it can only be good for us, but I think they underestimate ORSoC based on those crude assumed figures. The only thing Sam would commit to, and if anyone who was also there could back me up on this it would be useful, as I pressed and pressed for concrete info, was; 28nm, at least 250gh/s in Juliter (Saturn and Mercury weren't an option then), less than 1000w, and Jupiter. Aside from this they have kept their hand close to their chest with good reason.

I want to say I have no personal grievance directed towards Hashfast, as any negativity assumes unequivocal allegiance as a fan-boy, or if I bring up consumer protection and ccards, I'm declared a rogue agent provoceteur for Visa or Mcard, but simply my thoughts are this;

There is no movement and no attempt to offer any liability to any kind of third party willing to act as a guarantor in case of dispute to their claims and promises made. This option IS available to ALL western companies (100% in the US) as long as you satisfy their requirements. Crikey Butterfly Labs have for long enough. Terrahash have, Cointerra claim they will, and obviously KnC have. The real issue here is accountability and the responsibility of standing by your word. Hashfast aren't.

The tale of wanting to help the Bitcoin economy grow by increasing liquidity in the marginal amounts they exchange for dollars in comparison to the entire economy, is just that, a tale. This makes no impact, worse it takes the funds from the Bitcoin economy, reason being all their suppliers are paid in dollars. If they convinced all their suppliers to accept Bitcoin, that would be impressive, but they haven't.

Fact is HF are determined not to be held accountable for their originally promises October delivery. They have reworded their sales page to 'anticipated delivery'. They have reworded their refund offer only to be within a finite two weeks in January, long after funds will have been spent, and without a third party issuing bank to secure customer funds, a refund will be impossible. This makes me wonder how they intend to secure fabrication in 6 weeks? They cannot secure fabrication without funds in place, which they are still attempting to raise unsecured. They will require a rocket run, or a hot lot, and as far as I'm aware TSMC has no space left a ail able in that time. If they know this, and are basing their entire strategy around this, then they already know October is out of the question, as may well be November, hence no refund until January, when they will be low on funds.

Their ROI protection thing, means they will throw a chip your way, but you will be responsible for getting it assembled and online as they will be busy completing their backlog of orders. So you will get a chip, and have to send it off to be assembled with a third party, that will be in another queue and sent to you when ready. Time being of the essence and this being of additional cost to you, once you finally receive the additional product, the lions share of profit will have been made. In fact the biggest reward for all concerned will be mined between now and Christmas.

Then there is the warranty which is dismissed due to customers wishing to overclock. All chips from all manufacturers will be played with at somepoint. There is a well known variation of warranty used by vendors thought the globe with the majority of products made, it's called a 'manufacturers warranty', and within it manufacturers stipulate tolerances by which their product has acceptable warranty. You invalidate that, then it becomes your responsibility. To dismiss this as if their product is to be the only chip to be clocked is laughable. Butterfly Labs Jalepenos are running happily now at 8gh/s.

To me this looks like a gamble to raise funds ahead of Cointerra and X-Crowd without third party consumer protection mitigating liability so they can gamble customers funds on having a product ready by Christmas.  They already know now if they can fabricate a chip by October, and the behaviour in ensuring they cannot be held accountable for that suggests otherwise. Their alterations of terms and conditions after purchases have been made also suggests this maybe the case, which would be deceitful, if true. But by which point you will have committed yourself and be unable to recoup the funds spent. For sure there is no 100% refund, there is no way, they will have spent funds and have no one backing up the consumer. The safest play for us miners would be for ether to accept Paypal., because in 45 days by now, it would be October and you would know for sure if they could fabricate the chips, by when and when realisitcally to accept delivery. If by that point it looks untenable, you could simply ask Paypal for a refund and walk away. They don't want you to have this option available, they want to lock you in till Jan for a reason.

I was really looking forward to Cointerra's offering, I can only hope that's not been their last word on the subject as currently it's far from compelling for Dec/Jan delivery. That and despite promising Paypal, they prices at above $10k, Paypal's cut-off which indicated they never realistically looked into Paypal's requirements for sale.

In reality KnC should clear a chunk if not all ROI before Hashfast have a product. It doesn't not matter if they have a more efficient chip powerwise, if they are later than the end of October, by which point KnC should have cleared their backlog and potentially could sell in stock, competing on price as their NRE will have been recovered, offering a better deal to all. Will have to see...

Just got my full Jupiter refund.  Took about a week or so?

I was so late in the game (October delivery) that I doubt my pullout will have any effect on the rest of your orders.  I'll keep my eyes on the difficulty over the next several weeks as October approaches and will consider re-investing in the future.  For now my August BitFury will suffice.

just received mail from KNC asking them, when they will deliver my order:


dunno, what is "coming weeks", but now I'm pretty sure, that I will not receive my miner in first half of September.

Does Hashfast?  Does Cointerra?
Ask that Q next week. Pcb samples are most likely being tested as we speak. (With chip on board) Aug 14 news states we will be seeing PCB samples by the end of the week; but you probably knew that.
It's almost as if trolls were hired by BFL & Avalon, or even Hashfast to ask such Q's. Especially since Hashfast's "Why we're better than KNC" post. They couldn't even pre-sell 550 units at $8 a g/h.
Did you really NOT read the thread?... And decided to ask that Q in order to make a decision? I don't believe that to be true. So, with that in mind, the only logical reason one would ask that here is to stir the pot; and I can't express how much I really love that.
BTW, Orsoc & Kennmar & Cole AB have plenty of asics already delivered.(KNCminers parent companies) But you probably knew that too.

I said that you dont know if they have one, and you dont.   

You are joking or simply retarded? They have no working ASIC miner, which is the answer to seljo. If they did, they'd show it. I think even Bitcoinorama will agree on this.

If they would have packaged ASIC samples or even bare dies in hands (not even working), they would had published pictures for sure.
Come on, they even published pictures of the case and fans, which are absolute uncritical parts of their miner!

Chips are at the foundry and when they are done - that's when you know if they got a working ASIC or not. It could turn out to be duds but yeah.

No samples/prototype ASIC or whatever until the chips are done - and I'm fairly sure they'll tell you when they're done.

No one has even a working sample or a pre-production model?  With difficulties climbing like crazy, by the time it is out, I guess it will take forever to ROI

you dont know if they have a working ASIC.

FPGA - yes.
ASIC - no.

300 pages and I have only one question. Is there a single working KNC miner in existence? 

Amy’s estimation of the KnC die size was very unprofessional. There is no imperative relation between the size of the flip-chip package and the actual die size.

KnC has 192 unrolled hash cores, probably running at about 520 MHz to realize 100 GH/s in sum. To take advantage of the technology scaling the area of such a core should be around 0.5 mm2, which results in a KnC die size of about 100 mm2. If you multiply this by 4 there would be finally no big difference to a Hashfast chip. Which would not surprise me, because in the 28nm league only professionals play, who know how to use the full potential of this technology node.

900 mm2 are absolute ridiculous. With so much area one could realize 192 unrolled hash cores in 65nm!

I also understand that KnC keeps the real die size secret, because this could be used to calculate their marginal costs, which would be very usefull to know in the upcoming price war. But anyway, thanks Amy for the Hashfast strip!

This thread is beginning to suck, a lot like Christmas...

C'mon September, and then October!

+1

By then there will only be 500+ pages of trolling, opinions, accusations, analogies, and completely false statements and explanations made by those least qualified to make them.
Probably best to start a new thread IMHO. Any real information is greatly appreciated.

Hopefully by then we won't need to be posted.

Just wanted to say that I'm flying to Sweden next week to meet with the team over there. Will keep you guys posted.