BitcoinOrama Report on the KnCminer/OrSoC Open-day Mon 10/06/13 (Stockholm)

[KS]

and what is avalon doing?....they sale millions hundreds of thousand of avalon chips....it is the same thing

FTFY

[Bitcoinorama]

Marcus: We have done designs that are much more complex. One with 187 clock domains, that's the...that's the most tough design. I...hope that we will never see such a design again, so yes we have done more complex designs.

Me: What Was that?

Marcus: That was a hardcopy. 40Nm hardcopy.

Me: But what was that for?

Marcus: Ah that's a confidential large customer.

Is there any way I could ask them confidentially a question about that design? I know one that might just match that description of a previous -- likely a 65nm design -- that would benefit of being shrunk to 40nm. If it is really so, it would drastically change my view on the whole OrSoc, and I would have to say that they are grossly under-appreciated. I meant, you have a good relationship with them and this question really burns my mind.

And PLEASE, it is 40 nm as in nano-metres, not Newton-metres, oh my!!!!

Crikey, I think I've just been schooled by George Takei!!...

But you are right, I was used to the capital G in giga hash so accidentally gave nano the Newton N!

Feel free to email Rampion, if it doesn't jeapordise IP, they will answer straight.

I'm having a mare trying to upload video by the way. I can watch it back on the iPhone, fine. I can copy it from the iPhone, fine. The main 5 min file is around 900 megs as a .Mov, I have another of arounf 350 megs. Both my Mac and YouTube claim they don't recognise the file format!?

I have to goto work soon, but any ideas? Presumably I need to convert, is there a native format YouTube works with?? I'm surprised the Mac won't play it, .mov is apple after all. iPhone is the newer of the two though...

[KS]

Marcus: We have done designs that are much more complex. One with 187 clock domains, that's the...that's the most tough design. I...hope that we will never see such a design again, so yes we have done more complex designs.

Me: What Was that?

Marcus: That was a hardcopy. 40Nm hardcopy.

Me: But what was that for?

Marcus: Ah that's a confidential large customer.

Is there any way I could ask them confidentially a question about that design? I know one that might just match that description of a previous -- likely a 65nm design -- that would benefit of being shrunk to 40nm. If it is really so, it would drastically change my view on the whole OrSoc, and I would have to say that they are grossly under-appreciated. I meant, you have a good relationship with them and this question really burns my mind.

And PLEASE, it is 40 nm as in nano-metres, not Newton-metres, oh my!!!!

Crikey, I think I've just been schooled by George Takei!!...

But you are right, I was used to the capital G in giga hash so accidentally gave nano the Newton N!

Feel free to email Rampion, if it doesn't jeapordise IP, they will answer straight.

I'm having a mare trying to upload video by the way. I can watch it back on the iPhone, fine. I can copy it from the iPhone, fine. The main 5 min file is around 900 megs as a .Mov, I have another of arounf 350 megs. Both my Mac and YouTube claim they don't recognise the file format!?

I have to goto work soon, but any ideas? Presumably I need to convert, is there a native format YouTube works with?? I'm surprised the Mac won't play it, .mov is apple after all. iPhone is the newer of the two though...

convert to flash (flv)

[idee2013]

+1

FLV should be the right format for such i big file

[Bitcoinorama]

Ok, so I think it was a conflict between the formatting of the USB thumb drive not accepting the video files. The Mac is a Macbook Air and there just isn't enough space for me to play with 1.5 gigs of video currently.

So I've taken the iPhone to another PC (not mine) and the files play fine. They uploaded to YouTube as .MOV fine.

Daggeteo wanted to check I hadn't recorded anything sensitive from his pool data, so he gets first view. Personally I see nothing amiss. I almost stumble in on him typing his password which was amusing, aside from that there isn't an issue. He's briefly on screen for a millisecond, so his choice...nothing distressing or embarrassing.

Getting audio files off the iPhone is another headache entirely.

So lucky Daggeteo recorded audio separately. Mine is clearer, but anything over 8 minutes I cannot email. In iTunes I can see two out of the three voice memo in the 'on my iPhone' tab, click sync voice mails, nothing. Apprehensive to try pairing this with another computer as everyone I know has experienced total loss of all music and playlists doing this. Not risking loosing the audio. Incidentally the file that doesn't show on the computer, all though present on my iPhone, doesn't play on my iPhone. It was about 40 minutes long and it was the recording that stopped unannounced during the open-day. The other two files play uninterrupted. So again, lucky Daggeteo had a recording device present! There are additional bits on mine Daggeteo's doesn't have, but I need to have my iPhone plugged into a paired device to transcript. Joy...

This faffing around has cost me time, and is the reason there's been some delay. I have to head out to work now, if Daggeteo can give me a heads up by tomorrow. I shall share the video...

For what it's worth Apple has always been problematic with sharing large recordings made on iOS devices, they really need to address this!

On the plus side...it's in 1080p!!

[mandelmus]

Just think: if KnC sells 1,500 Jupiters and delivers them quickly, when added to the rest of the network each one will probably produce aprox. 1BTC or less per day, and ROI will be unlikely. If they sell +3,000 Jupiters (which seems likely they will do), when they arrive each Jupiter will produce 0,5BTC per day or less, and ROI won't never happen.

1/2 a BTC per day?!  That's what worries me, especially for those of us mining without Jupiters and Saturns.

[Rampion]

Just think: if KnC sells 1,500 Jupiters and delivers them quickly, when added to the rest of the network each one will probably produce aprox. 1BTC or less per day, and ROI will be unlikely. If they sell +3,000 Jupiters (which seems likely they will do), when they arrive each Jupiter will produce 0,5BTC per day or less, and ROI won't never happen.

1/2 a BTC per day?!  That's what worries me, especially for those of us mining without Jupiters and Saturns.

ASIC mining is a very particular beast with very special characteristics. If you want your customers to achieve ROI, you cannot just flood the network as quick as possible. That's why the "first 500 batch get their unit first" approach makes some sense, while the "we can push hundreds of units daily and everybody will receive them almost at the same time" is a kinda worring approach from a preorder customer point of view.

I expect a 2011-like situation, when we had a "mining bubble" that deflated because it wasn't profitable to mine any more. I see the very same situation approaching very fast, the absolute winners have been the ASICminer shareholders and Avalon batch #1 customers, batch #2 will ROI and #3 will breakeven, and I'm afraid that the rest (including BFL customers) will be the "bagholders".

I may be mistaken, but I would appreciate some figures from KnC regarding how many THs they plan to sell, and how fast. As I said earlier this is a very "special" business, if they sell too many units too fast their customers won't achieve ROI unless they hold their mined coins till the exchange rate goes x10, and in that case their profit would have been bigger if they just bought BTC instead of buying KnC miners.

Making closed and limited batches is the best way to try to protect your customers investment in this business.

[Bagpipe]

flv is a container, not a codec.
.mov is also a container only and inside it can be mjpeg (motion jpeg a series of photos) or some other codec.

There is a utility on sourceforge named mov2mpeg or something like that...
no, I believe it had the words mov, 2, mp4 in the name... I can not tell you as it is on a crashed computer.


As for the KnC miner numbers, they totally, utterly had a plan of flooding the market with something that may be a collective 30% saturation. The numbers sold by the preorders were far less than what they expected to sell and the expenses for this project will run into 4 to 6 million USD when we count their engineering work too. So selling KnC miners at least to a break-even point and NOT ruining the BTC network seems like an impossible task.

[mandelmus]

As for the KnC miner numbers, they totally, utterly had a plan of flooding the market with something that may be a collective 30% saturation. The numbers sold by the preorders were far less than what they expected to sell ...

How do you know it was far less than they expected?

... and the expenses for this project will run into 4 to 6 million USD when we count their engineering work too.

I'm sure they have surpassed 7 million USD by this point.  Does anyone have the actual sales numbers yet?

So selling KnC miners at least to a break-even point and NOT ruining the BTC network seems like an impossible task.

So, you are thinking that KnC miners will ruin the BTC network? 

There are so many factors to consider:  Will KnC be able to deliver by September?  Who else will flood the market with ASICs before KnC gets their products delivered?  Where will the exchange-to-BTC rate be in September ... what if the exchange rate drops back down to 5 USD per BTC?  How will this flood of ASICs affect us little guys who are mining for only pennies a day?

On the other hand, could this be a really good thing for the Bitcoin economy?  Will the Bitcoin economy become stronger, more effective, efficient and resilient, with this increased flooding of ASICs on the market? 

[johnyj]

Just think: if KnC sells 1,500 Jupiters and delivers them quickly, when added to the rest of the network each one will probably produce aprox. 1BTC or less per day, and ROI will be unlikely. If they sell +3,000 Jupiters (which seems likely they will do), when they arrive each Jupiter will produce 0,5BTC per day or less, and ROI won't never happen.

1/2 a BTC per day?!  That's what worries me, especially for those of us mining without Jupiters and Saturns.

ASIC mining is a very particular beast with very special characteristics. If you want your customers to achieve ROI, you cannot just flood the network as quick as possible. That's why the "first 500 batch get their unit first" approach makes some sense, while the "we can push hundreds of units daily and everybody will receive them almost at the same time" is a kinda worring approach from a preorder customer point of view.

I expect a 2011-like situation, when we had a "mining bubble" that deflated because it wasn't profitable to mine any more. I see the very same situation approaching very fast, the absolute winners have been the ASICminer shareholders and Avalon batch #1 customers, batch #2 will ROI and #3 will breakeven, and I'm afraid that the rest (including BFL customers) will be the "bagholders".

I may be mistaken, but I would appreciate some figures from KnC regarding how many THs they plan to sell, and how fast. As I said earlier this is a very "special" business, if they sell too many units too fast their customers won't achieve ROI unless they hold their mined coins till the exchange rate goes x10, and in that case their profit would have been bigger if they just bought BTC instead of buying KnC miners.

Making closed and limited batches is the best way to try to protect your customers investment in this business.

Even current ASIC manufacturers make some organization like OPEC, due to there are hidden players developing ASICs mining for themselves, that does not work either. One of the most important mission of ASIC miner manufacturers is to promote the bitcoin-only sale to support the price of bitcoin (hence their own income)

[rizzman]

Even current ASIC manufacturers make some organization like OPEC, due to there are hidden players developing ASICs mining for themselves, that does not work either. One of the most important mission of ASIC miner manufacturers is to promote the bitcoin-only sale to support the price of bitcoin (hence their own income)

This is simply not true... and even if it was, it would be disastrous to bitcoin long term considering all you have at that point is an exclusive club where the only persons who have access to the tools of the trade are those that had got into the mining market at ground level and have a hoard of bitcoin. Meanwhile these companies producing their mining products are reliant on third-party vendors to exchange their bitcoin to pay the vendors that only accept local currency for the raw material needed to make the finished products. This is NOT good for business.

This practice is just not tenable and you can see that in the moves that most of the major ASIC manufacturers are making towards including PayPal's shitty service into their payment options (ie. BFL, Terrahash, KNCMiner) because they know that people that weren't in the game before are already scared as it is moving into a highly volatile new currency let alone, the promise of generating it, and will need a comfortable entry point into the market in order to have any chance of growth.

Sure you can counter argue the use of Mt.Gox and other exchanges are there to serve as intermediaries but the reality is people are lazy, scared, and resistant to change... And the easier you make for them to enter into the market, the more sustainable the growth and confidence in it will be long term. The companies know/are learning this and is one of the best chances that Bitcoin has to succeed.

[mandelmus]

... Meanwhile these companies producing their mining products are reliant on third-party vendors to exchange their bitcoin to pay the vendors that only accept local currency for the raw material needed to make the finished products. ...

Ya, that's a really good point. 

All these ASIC developers/manufacturers are asking for and accepting payment in Bitcoin, but they still have to convert those bitcoins into fiat currency to pay their vendors.  Theoretically, if everyone paid KnCMiner in Bitcoins -- approximately $7 million so far -- shouldn't we start seeing some pretty large BTC sell orders on MtGox soon?

[Korbman]

Theoretically, if everyone paid KnCMiner in Bitcoins -- approximately $7 million so far -- shouldn't we start seeing some pretty large BTC sell orders on MtGox soon?

ASIC manufacturers, namely BFL and KnC (though also Avalon at one point), all utilized BitPay (or similar service) for Bitcoin based transactions. What this means is that those companies never saw/received a coin...it was instantly converted to USD (or whatever other currency) at the market rate at the time of purchase.
Bitcoins --> BitPay --> converted to cash --> Manufacturer is paid (and BitPay takes a cut).

[hardpick]

Why is everybody so happy that they can produce and deliver so many units per day?

Don't you realize this is a race, and the best way to maximize ROI for customers is to sell batches, as Avalon did?

Just think: if KnC sells 1,500 Jupiters and delivers them quickly, when added to the rest of the network each one will probably produce aprox. 1BTC or less per day, and ROI will be unlikely. If they sell +3,000 Jupiters (which seems likely they will do), when they arrive each Jupiter will produce 0,5BTC per day or less, and ROI won't never happen.

Honestly, it's pretty scary to see them pushing out those numbers. Each additional Jupiter they sell removes value from the previous one. Kinda the same problem we thought BFL would have, but the reality ended up being that they are so slow delivering that actually their very first customers may break/even with their units despite the delays.

and what is avalon doing?....they sale millions of avalon chips....it is the same thing

BFL sells chips, too. But only 100,000 https://products.butterflylabs.com/65nm-asic-bitcoin-mining-chip.html

any idea what the TH will be in November / dec  2013 ?

is there a link someware to bitcoin hashrate  by end of year predictions ?


my guess:
bfl current orders  ~ 400 Th -- Avalon chips ~ 300 TH --- KnC ~350 TH  ---- Asicminer ~ 100 TH --- BFL chips~ 400 TH
+ Current network ~ 136 TH  = about 1686 TH

[innovation]

Why is everybody so happy that they can produce and deliver so many units per day?

Don't you realize this is a race, and the best way to maximize ROI for customers is to sell batches, as Avalon did?

Just think: if KnC sells 1,500 Jupiters and delivers them quickly, when added to the rest of the network each one will probably produce aprox. 1BTC or less per day, and ROI will be unlikely. If they sell +3,000 Jupiters (which seems likely they will do), when they arrive each Jupiter will produce 0,5BTC per day or less, and ROI won't never happen.

Honestly, it's pretty scary to see them pushing out those numbers. Each additional Jupiter they sell removes value from the previous one. Kinda the same problem we thought BFL would have, but the reality ended up being that they are so slow delivering that actually their very first customers may break/even with their units despite the delays.

and what is avalon doing?....they sale millions of avalon chips....it is the same thing

BFL sells chips, too. But only 100,000 https://products.butterflylabs.com/65nm-asic-bitcoin-mining-chip.html

any idea what the TH will be in November / dec  2013 ?

is there a link someware to bitcoin hashrate  by end of year predictions ?


my guess:
bfl current orders  ~ 400 Th -- Avalon chips ~ 300 TH --- KnC ~350 TH  ---- Asicminer ~ 100 TH --- BFL chips~ 400 TH
+ Current network ~ 136 TH  = about 1686 TH

I think Acisminer will be 500+TH.

[dan99]

I think some of this comments can best be on mining speculation board/thread

[Rampion]

Why is everybody so happy that they can produce and deliver so many units per day?

Don't you realize this is a race, and the best way to maximize ROI for customers is to sell batches, as Avalon did?

Just think: if KnC sells 1,500 Jupiters and delivers them quickly, when added to the rest of the network each one will probably produce aprox. 1BTC or less per day, and ROI will be unlikely. If they sell +3,000 Jupiters (which seems likely they will do), when they arrive each Jupiter will produce 0,5BTC per day or less, and ROI won't never happen.

Honestly, it's pretty scary to see them pushing out those numbers. Each additional Jupiter they sell removes value from the previous one. Kinda the same problem we thought BFL would have, but the reality ended up being that they are so slow delivering that actually their very first customers may break/even with their units despite the delays.

and what is avalon doing?....they sale millions of avalon chips....it is the same thing

BFL sells chips, too. But only 100,000 https://products.butterflylabs.com/65nm-asic-bitcoin-mining-chip.html

any idea what the TH will be in November / dec  2013 ?

is there a link someware to bitcoin hashrate  by end of year predictions ?


my guess:
bfl current orders  ~ 400 Th -- Avalon chips ~ 300 TH --- KnC ~350 TH  ---- Asicminer ~ 100 TH --- BFL chips~ 400 TH
+ Current network ~ 136 TH  = about 1686 TH

With 1686TH/s on the network a 350GHs Jupiter will generate 0.7BTC per daily. That would have been an awful investment.

[KS]

With 1686TH/s on the network a 350GHs Jupiter will generate 0.7BTC per daily. That would have been an awful investment.

That would give you a ROI of 3-4 months at current exchange rate, that's pretty normal. However, there is no reason the difficulty won't keep on rising and if it rises at a continued 25-30% rate, they'll never make ROI (but neither will anyone buying in September/October unless there is a new generation of ASICs with better performance and energy consumption).

[fragout]

Well there will be a small decrease from gpu miners jumping ship and the low hashing ascis. Also the other sha coins will distrubute the load a bit, but the only way any1 is going to make a few $ is if the price of btc increases which is the big unknown in the whole equation.

[idee2013]

Well there will be a small decrease from gpu miners jumping ship and the low hashing ascis. Also the other sha coins will distrubute the load a bit, but the only way any1 is going to make a few $ is if the price of btc increases which is the big unknown in the whole equation.

+1