HashFast announces specs for new ASIC: 400GH/s

[crumbs]

...
I didn't go to the open house but I did visit them personally before that. I also signed an NDA and was happy to do so because they gave more information out than just about any company currently out there without physical chips.

You can't fault them for that, right now as long as everyone at the open house can agree that the specs they have shown are valid, that this company is a legit company, and that they are still on good standing to deliver when they say they will, what more can you ask for. ...

An ASIC company with zero track record & zero cred holds an open house where people are turned away, and the select few who make it in are made to sign NDAs.  I think i got all i need

Edit: An interesting IPO being offered here by DeadTerra (not to be confused with Cointerra), involving HashFast (not to be confused with Fast Hash).  Any comments, considering there's talk of a deal with first HashFast devices going exclusively to DeadTerra?

You can only fit so many people in a building, even you should have the capability to understand this.
They were accepting visits before this "open house" and you could have chosen not to sign an NDA, you would just receive less information.

I'm sorry if you're used to being being told to GTFO when you appear at an open house.  Never happened to me.  If you are also used to signing NDAs when you come to an open house, i suspect you're being trolled.

There isn't really talk of a deal, DeadTerra made it pretty straight forward in his IPO Thread.

Q:
Have HashFast said they will only sell chips to IceDrill ?

A:
We have a deal with HashFast that guarantees that we will get the chips from the first batch produced.
Once the chips are off from the fab they will be sent straight to us for assembly.

It pains me when people can't use basic logic to dictate why people do what they do.

It pains me to see that you're discussing what it pains you to see, instead of topics pertinent to the thread -- deals which will affect investors.  I'm not as concerned about your pains as i'd like to be, so let's stay on topic.

[crumbs]

...
I didn't go to the open house but I did visit them personally before that. I also signed an NDA and was happy to do so because they gave more information out than just about any company currently out there without physical chips.

You can't fault them for that, right now as long as everyone at the open house can agree that the specs they have shown are valid, that this company is a legit company, and that they are still on good standing to deliver when they say they will, what more can you ask for. ...

An ASIC company with zero track record & zero cred holds an open house where people are turned away, and the select few who make it in are made to sign NDAs.  I think i got all i need

Edit: An interesting IPO being offered here by DeadTerra (not to be confused with Cointerra), involving HashFast (not to be confused with Fast Hash).  Any comments, considering there's talk of a deal with first HashFast devices going exclusively to DeadTerra?

You can only fit so many people in a building, even you should have the capability to understand this.
They were accepting visits before this "open house" and you could have chosen not to sign an NDA, you would just receive less information.

I'm sorry if you're used to being being told to GTFO when you appear at an open house.  Never happened to me.  If you are also used to signing NDAs when you come to an open house, i suspect you're being trolled.

There isn't really talk of a deal, DeadTerra made it pretty straight forward in his IPO Thread.

Q:
Have HashFast said they will only sell chips to IceDrill ?

A:
We have a deal with HashFast that guarantees that we will get the chips from the first batch produced.
Once the chips are off from the fab they will be sent straight to us for assembly.

It pains me when people can't use basic logic to dictate why people do what they do.

It pains me to see that you're discussing what it pains you to see, instead of topics pertinent to the thread -- deals which will affect investors.  I'm not as concerned about your pains as i'd like to be, so let's stay on topic.

You say there isn't going to be a 28nm ASIC in other threads,

Refer to yourself as a "We" with CoinTerra,

then are directly concerned with investors of HashFast now?

If anyone believed in a word you said before, I'm sure they have quickly come around now.

As i've told you before, i'm behind all of it -- Cointerra, Deadterra, TerraHash (whoops -- not 'till next week) Fast Hash, Hashcoin, and Miracle Grow Investments.  I'm known as The Hamburglar IRL.  You got me

[bitcoinarnold]

NDA to an open day? LOL fuck that

[Ytterbium]

Making everyone sign an NDA is pretty stupid.  I requested to go to the open house and was told: 'no' we're full.  Not a huge fan of these guys right now.  We know that they had an open house.  However, we have no clue what was discussed.  So, what was the point of it?

It's kind of annoying.

On the other hand, their main public competition right now is Cointerra, it's probably something of a game of chicken with them, they don't want to put out specs before Cointerra does if it means Cointerra can try to one up them.

Anyway, as far as I'm concerned the hashrate is going to be going up way to quickly to bother buying any more hashpower at this point in time.  We'll have to see where things are in December, and what the prices are.

I'm a little worried people aren't really doing their research or thinking clearly are going to order a ton of product they'll never ROI off of, and at the same time destroy profitability for everyone else as well.

Thanks to everyone that came out for our Office visit in San Jose this morning!
(hopefully some of them will post about it

John

But you made them sign an NDA so they can't talk about it.  Congratulations

Heh.

It pains me to see that you're discussing what it pains you to see, instead of topics pertinent to the thread -- deals which will affect investors.  I'm not as concerned about your pains as i'd like to be, so let's stay on topic.

You say there isn't going to be a 28nm ASIC in other threads,

Refer to yourself as a "We" with CoinTerra,

then are directly concerned with investors of HashFast now?

If anyone believed in a word you said before, I'm sure they have quickly come around now.

Crumbs is a troll, dude.  Earlier he was saying that HashFast hadn't even indicated that it ever planned to produce a chip.  He's not trying to have a discussion, he's trying to bait people into jumping through his pseudo-logical 'arguments' that actually don't make any sense at all.

Just click the ignore button.

[Flashman]

Anyyyway, I guess anything seen before the NDA was signed is fair game....



So do they have a nice parking lot? Are the lobby chairs comfy?

[Loredo]

Anyyyway, I guess anything seen before the NDA was signed is fair game....



So do they have a nice parking lot? Are the lobby chairs comfy?

I hope nobody parked in the space that says "Lumbergh" on it.

And I thought it was going to be a quiet week.  Instead, it's turning into the wrong week to quit sniffing glue.

[crumbs]

Anyyyway, I guess anything seen before the NDA was signed is fair game....



So do they have a nice parking lot? Are the lobby chairs comfy?

The NDA was lowered in a basket, on a string.  
The pen was in the same basket.  
The basket had a note pinned to it: "It will rub the lotion on its skin or else it will get the hose. It will sign or GTFO. Do it."

I'm afraid no lobby descriptions for you.

[Ytterbium]

If you do the math on IceDrill shares, it comes out to about $14/Gh/s starting in November.

50 million shares at 0.0014btc each.
500Th/s worth of HashFast chips.

So, one share costs you $0.14, and gets you 10Mh/s, or $14 for 100 shares and 1Gh/s.

So, either HashFast is charging $14/Gh, or IceDrill is making a lot more profit off this deal then they're letting on. That's the problem with this NDA stuff, it makes it hard to evaluate various deals.

On the other hand Labcoin, which has their own chip, is offering ~520Th/s by December (Starting from 50Th in September). That comes out to about $3.50/Gh/s in December and about $36/Gh/s in September. Plus with labcoin you actually get a cut of the chip sales, instead of simply seeing the difficulty going up.

[Loredo]

Edit: An interesting IPO being offered here by DeadTerra (not to be confused with Cointerra), involving HashFast (not to be confused with Fast Hash).  Any comments, considering there's talk of a deal with first HashFast devices going exclusively to DeadTerra?

Thanks for posting this link, crumbs.  DeadTerra has a good reputation and I am glad he is putting together the IceDrill IPO based on HashFast.  Right now there are few ways to get into the game besides buying actual mining rigs/chips.  I will definitely be looking into DeadTerra's IPO.  I suspect everyone will dump their shares of other mining operations to buy HashFast's shares since their claims, if true, would dominate the competition.

You see why pre-orders are such bullshit, both for the buyers and sellers?  The whole network has only touched 400 this week, and as I write this is humming along at around 340.  

So now, here comes a guy who says, in effect: "Uh, yeah, were coming with 500 in November, and we're going to hold on to some chunk of the total thereafter.  You all are welcome to buy in to 60% of our action, because if there aren't some small stakeholders, and there's only six of us in the whole network, it won't be good."

So, then ASICMiner has to increase.  And buzzDave has to increase.  And BFL has to smuggle a couple more minirigs out the back door.  And KnC has to redefine its own hosting, and the hosting deals they've already made. And God knows what goes on in China with those now-commoditized 130nm chips.  Oh, yeah, and I forgot CoinTerra.   We don't know what they're going to do yet, I guess.

The point is, any growth rate you put on this stuff can be -no, will be- black swanned.  You wanna buy fixed hash, delivered to you at an uncertain time point in the future?  Bend over.

[Ytterbium]

You see why pre-orders are such bullshit, both for the buyers and sellers?  The whole network has only touched 400 this week, and as I write this is humming along at around 340.  

It actually got close to hitting 500: https://i.imgur.com/dBEVHmK.png

So now, here comes a guy who says, in effect: "Uh, yeah, were coming with 500 in November, and we're going to hold on to some chunck of the total thereafter.  You all are welcome to buy in to 60% of our action, because if there aren't some small stakeholders, and there's only six of us in the whole network, it won't be good."

So, then ASICMiner has to increase.  And buzzDave has to increase.  And BFL has to smuggle a couple more minirigs out the back door.  And KnC has to redefine its own hosting, and the hosting deals they've already made. And God knows what goes on in China with those now-commoditized 130nm chips.

The point is, any growth rate you put on this stuff can be -no, will be- black swanned.  You wanna buy fixed hash, delivered to you at an uncertain time point in the future?  Bend over.

Someone has to do a pre-order, other way there is no way to get chips.  If you put in an order a the fab, it takes 3 months get your chips. That's a pre-order.

So the only way to produce ASICs is to crowd-fund through pre-orders or an IPO, or fund your venture privately which means only people with hundreds of thousands, or millions of dollars to burn can get involved in Mining.

This idea that some people have that a company is going to invest millions of their own dollars into designing and producing their own chips, then give them away for far less then what they'll earn on the network with same day shipping instead of mining with them themselves is a pipe dream.

No one is ever going to sell chips they have on hand and paid for for less then what they would expect to make in profit over whatever time horizon they're looking at.

Now, that cost will go down as the difficulty goes up - and it might get so high that chip companies won't even bother trying to mine, in which case chips will be very cheap, and mining will barely cover electrical costs.

[Flying Hellfish]

Edit: An interesting IPO being offered here by DeadTerra (not to be confused with Cointerra), involving HashFast (not to be confused with Fast Hash).  Any comments, considering there's talk of a deal with first HashFast devices going exclusively to DeadTerra?

Thanks for posting this link, crumbs.  DeadTerra has a good reputation and I am glad he is putting together the IceDrill IPO based on HashFast.  Right now there are few ways to get into the game besides buying actual mining rigs/chips.  I will definitely be looking into DeadTerra's IPO.  I suspect everyone will dump their shares of other mining operations to buy HashFast's shares since their claims, if true, would dominate the competition.

You see why pre-orders are such bullshit, both for the buyers and sellers?  The whole network has only touched 400 this week, and as I write this is humming along at around 340.  

So now, here comes a guy who says, in effect: "Uh, yeah, were coming with 500 in November, and we're going to hold on to some chunk of the total thereafter.  You all are welcome to buy in to 60% of our action, because if there aren't some small stakeholders, and there's only six of us in the whole network, it won't be good."

So, then ASICMiner has to increase.  And buzzDave has to increase.  And BFL has to smuggle a couple more minirigs out the back door.  And KnC has to redefine its own hosting, and the hosting deals they've already made. And God knows what goes on in China with those now-commoditized 130nm chips.  Oh, yeah, and I forgot CoinTerra.   We don't know what they're going to do yet, I guess.

The point is, any growth rate you put on this stuff can be -no, will be- black swanned.  You wanna buy fixed hash, delivered to you at an uncertain time point in the future?  Bend over.

Ever since becoming involved with mining I have wondered why so many people are so willing to fund someone else's business (and shoulder a majority of the risk) in return for a product they paid full price for.  In the real world there is a big difference between investors and customers (one funds the business and one buys goods or services), here in Deadwood we have investomers.  I'll never get it that's for darn sure!

[Loredo]

This idea that some people have that a company is going to invest millions of their own dollars into designing and producing their own chips, then give them away for far less then what they'll earn on the network with same day shipping instead of mining with them themselves is a pipe dream.

You're assigning too little weight to an essential economic fact, and one that has influenced the way venture investment has proceeded in bitcoin, and will continue to proceed.  

That is: bitcoins are not money.  They have to be converted to money.  A big miner has to convert a quantity which grows continuously, and do so without the benefit of a real futures market, or the ability to short sell in any meaningful size, and without significant transactional - that is, non financial - demand.  That's not easy.  

How do you convert them to money?  Well, one way would be to make machines, and sell them - for money - to people who mine with them.  What they do with the bitcoins is their problem.  

Do you know the name Meyer Guggenheim?  He was a bright young man who went out to the gold rush in California, and quickly figured out that to make the big money, day in and day out, one should refine the gold, not mine it.  The Guggenheim fortune was and is one of largest in American history.

EDIT: In light of a court document that's making the rounds from the Pirate case, someone might take issue with my claim that bitcoins are not money.  In the sense I'm using the term here, the legal or even economic "money nature" of bitcoins is not important.  What's important is this:  Right now, you could sell 35million EUR for $1.33857, or buy 18million EUR for $1.33860, and that's just on the screen open in the background where I'm typing.  You can buy or sell 10 thousand ounces of gold for prices only a few cents per ounce different.  You can't do that -yet- in bitcoin. 

So, strictly, I shouldn't have said, bitcoins are not money.  I should have said bitcoins have poor fungibility.

[AfricanHunter]

What is your refund policy on preorders?

When will you announce actual specs/pricing/delivery dates?

[Ytterbium]

ASICs that utilize 130nm or 65nm cannot compete because more silicon is required to make these chips.  Plus, it generally takes more power and generates more heat.  Any company that is utilizing 65nm designs won't last long term because they simply cannot compete with 28nm.

That's not going to matter until most of the profitability in mining is gone and the value you get from mining is close to the cost of electricity.

Right now, the goal when buying mining equipment is to have it pay for itself in a few months, $/Gh/s is going to matter more then Gh/J.

Since the NRE costs are lower for higher densities, the cost of the chips may be lower in the short run, and only after costs are amortized will 28nm be cheaper then higher densities.

Obviously in the long run we'll be running 22, 20nm designs.

Interesting.  What I do know is that LabCoin uses 130nm and 65nm design versus HashFast's 28nm design.  So, basically LabCoin is working on first and second generation ASIC bitcoin miners while HashFast is already working on third generation.  It is possible long term LabCoin will not be able to compete with HashFast.  What this may mean is IceDrill may be able to upgrade and put more efficient 28nm units in faster while LabCoin will still be working on their 65nm design which will be rendered obsolete.

Well, the difference is LabCoin (And ASICminer, BTCGarden and ActiveMining) pays whatever the fab charges for it's chips, while IceDrill has to pay whatever HashFast wants to charge them.

Anyway, the thing to keep in mind is that unlike a CPU, you can scale these out in a nearly completely linear way, since the necessary bandwidth between chips is so low.

Using this profitabilty calculator an 82Gh/s, 110nm Avalon unit drawing 700w would make about $110/day until the next difficulty increase.

The power costs would be about $1.60, at $0.10/kWh.

That means in order for the unit to be unprofitable @ $0.10/kWh, the difficulty would need to be about 2.5 billion

And in order for the unit to be unprofitable at $0.05/kWh, the difficulty would have to be over billion.

So think about it: You get a 400Gh/s HashFast chip @ $14/Gh  That comes out to $5,600.  Let's assume this magical 28nm efficiency lets it run at 100W.

At 2.5 billion difficulty, you only make $8/day, and it would take almost two years to earn back your investment
At 5 billion difficulty, you only make $4/day and it takes you almost 4 years to earn back your investment!

See the problem?   While it's true that IceDrill might need to use less electricity then LabCoin today, that power cost is going to be far lower then the money they'll make though mining in the short run.

And on top of that, the money they'll make in the long run is hardly anything, compared to what they'll make in the short run.

The way things are going, I actually wouldn't be surprised to see a few TH difficulty in the early part of next year.

[dan99]

If you do the math on IceDrill shares, it comes out to about $14/Gh/s starting in November.

50 million shares at 0.0014btc each.
500Th/s worth of HashFast chips.

So, one share costs you $0.14, and gets you 10Mh/s, or $14 for 100 shares and 1Gh/s.

So, either HashFast is charging $14/Gh, or IceDrill is making a lot more profit off this deal then they're letting on. That's the problem with this NDA stuff, it makes it hard to evaluate various deals.

On the other hand Labcoin, which has their own chip, is offering ~520Th/s by December (Starting from 50Th in September). That comes out to about $3.50/Gh/s in December and about $36/Gh/s in September. Plus with labcoin you actually get a cut of the chip sales, instead of simply seeing the difficulty going up.

Interesting.  What I do know is that LabCoin uses 130nm and 65nm design versus HashFast's 28nm design.  So, basically LabCoin is working on first and second generation ASIC bitcoin miners while HashFast is already working on third generation.  It is possible long term LabCoin will not be able to compete with HashFast.  What this may mean is IceDrill may be able to upgrade and put more efficient 28nm units in faster while LabCoin will still be working on their 65nm design which will be rendered obsolete.

Timing and delivery time is important in this game. The faster the better. Labcoin has already done the tape out for the 130nm asic chip and now working on the 65nm tape out which they say should be this very month. Their vehicles may not have the luxury compares to a Chrysler 300 or Chevrolet Corvette Z06, but hey theirs can also get you from point A to point B.

[solex]

So think about it: You get a 400Gh/s HashFast chip @ $14/Gh  That comes out to $5,600.  Let's assume this magical 28nm efficiency lets it run at 100W.

At 2.5 billion difficulty, you only make $8/day, and it would take almost two years to earn back your investment
At 5 billion difficulty, you only make $4/day and it takes you almost 4 years to earn back your investment!

See the problem?   While it's true that IceDrill might need to use less electricity then LabCoin today, that power cost is going to be far lower then the money they'll make though mining in the short run.

And on top of that, the money they'll make in the long run is hardly anything, compared to what they'll make in the short run.

The way things are going, I actually wouldn't be surprised to see a few TH difficulty in the early part of next year.

Correct me if I am wrong but the "Conversion rate (USD/BTC)" value which is used in this site is a constant? http://www.bitcoinx.com/profit/
Since difficulty is projected to increase, and price has done the same for 4.5 years, then surely there should be a month-by-month projection of conversion rate too. A lot of mining might be unprofitable at $100 but profitable at $1000 which could easily be the case when difficulty is 2.5 billion!

[Ytterbium]

Correct me if I am wrong but the "Conversion rate (USD/BTC)" value which is used in this site is a constant? http://www.bitcoinx.com/profit/
Since difficulty is projected to increase, and price has done the same for 4.5 years, then surely there should be a month-by-month projection of conversion rate too. A lot of mining might be unprofitable at $100 but profitable at $1000 which could easily be the case when difficulty is 2.5 billion!

Sure... but if you think bitcoin is going to go up in value, you can just buy BTC today.

Or, you could just not spend the bitcoins you do have to buy a miner.  I'm sure the people who spent 3k BTC on BFL Mini-rigs certainly wish they'd done that!

[gateway]

So I also have faith in HashFast, and knowing John for a while I was able to talk to them quite a while a few weeks back, yes nda on my side blah blah blah as well sorry guys..

but I think these are the main things we are all concerned about (except for some idiot who just paid some 65k for a bfl mini rig) sorry couldn't help myself.

Time is money

Diff is going up and not down, I think the last block change it went up from 31 million to 37 millon, so give or take every 11 days or so we can expect this to grow.  So the days of selling maybe overpriced hardware is over, at least for the smart investors

If companies coming out towards the end of this year are still charging between 6k-10 or more for anywhere from 300-500Gh/s their just is NO ROI anymore. 

So yes the price of $$ per Gh/s will come down, and its important to get out quick for these companies they also need to make a profit as well but the time for the greedy companies is over..

anyhow its late im rambling , spent all freaking day load testing our servers and im spent.. but I know for sure you guys will like what HashFast will be offering

[Bitcoinorama]

Me in the office.

Hey! You got biscuits!!

[gateway]

Me in the office.

Hey! You got biscuits!!

I think those are jello shots next to him..