HashFast announces specs for new ASIC: 400GH/s

[dhenson]

I think we can rest assured there's no second chip taping out in a week.

Wow, nice 20/20 hindsight bro.  The reason nobody accepted the wager a year ago was that the success of the GN1 made production of the GN2 a non-trivial probability.

How au courant of you to perform a little victory jig on HF's grave at this late date.  Aren't you just tickled pink that a BTC company failed, leaving its customers in limbo?

As for the Zappa copypasta guy, if you are unfamiliar with and surprised by the quotidian industry-wide practice of separating IP from production LLCs, please put all 100% your BTC in a cold wallet and never buy or invest in anything.  Because you are a clueless, dewy-eyed, greenhorn N00B...   

[1714268434]

1714268434

[RoadStress]

When a risky high-tech Silicon Valley runs out of money and declares bankrupcy, it's a common and perfectly normal event.  Likewise for older and bigger firms.

No "prepackaged bankruptcy scam" would produce a chip like the GN1, which at well over 700GH/s is still by far the fastest BTC ASIC on the planet.

But lying about delivery dates sound like a "prepackaged scam"...How on earth do you promise November delivery when it was clear from the start that they couldn't meet it?

[Gyrsur]

I want to buy an used upgrade kit for my BabyJet. From where I'm able to get one?

http://hashfast.com/shop/baby-jet-upgrade-kit/

EDIT: My account is gone in the hashfast system --> "There is no user registered with that email address."

[Gyrsur]

Looks like the BitCrane T-110 is using HF boards:

http://www.bitcrane.com/down/T-110_en.pdf

http://www.cryptocoinsnews.com/news/review-bitcrane-t-110-1-1-ths-watercooled-bitcoin-asic-miner/2014/08/28

http://www.cryptocoinsnews.com/wp-content/uploads/2014/08/20140826_013915.jpg

Anybody know more about these guys?

HF did sell their Yoli Evo boards to this company above? And did not ship it to their customers?

http://hashfast.com/shop/yolievo/

[kikaha]

I talked to Bitcrane and asked this exact question.  They paid for the Yoli EVO design files and the chips and built boards in China.  So other than buying chips that should have been going towards customer orders, they appear to be legit.

Looks like the BitCrane T-110 is using HF boards:

http://www.bitcrane.com/down/T-110_en.pdf

http://www.cryptocoinsnews.com/news/review-bitcrane-t-110-1-1-ths-watercooled-bitcoin-asic-miner/2014/08/28

http://www.cryptocoinsnews.com/wp-content/uploads/2014/08/20140826_013915.jpg

Anybody know more about these guys?

http://www.bitcrane.com/images/detail_06.jpg

HF did sell their Yoli Evo boards to this company above? And did not ship it to their customers?

http://hashfast.com/shop/yolievo/

[2112]

As for the Zappa copypasta guy, if you are unfamiliar with and surprised by the quotidian industry-wide practice of separating IP from production LLCs, please put all 100% your BTC in a cold wallet and never buy or invest in anything.  Because you are a clueless, dewy-eyed, greenhorn N00B...    

The mention of IP (as Intellectual Property) is worth replying. Because there is none.

SHA256 brute forcer is such a trivial circuit, that anyone can design and make it. For students this would be a semester project, for experienced engineer with access to the appropriate toolchain this would be a couple of weekends. All the value is in the execution: people who have the "know how" and "from manufacturing to fulfillment" channel. Once those people left Hashfast and went back to Uniquify the value of the "IP" went to near zero, less than the value of the software licenses of the tools used to produce it.

Because there's no real, separable, intellectual property it would be better to account for it as a
http://en.wikipedia.org/wiki/Goodwill_(accounting) . There are no patents, trademarks, brand names, etc. that would have a value/worth to a different organization.

Bitcoin gives the users the power to disintermediate the conventional banks and classical accounting. This just means that the work done (expensively) by the bank or accountancy can be done (cheaper) by the end user. But not doing that work at all converts the investors into gamblers.

Conventional finance uses letters of credit to buy/manufacture custom equipment. It is similar to escrow, with the difference that LoC can be used as a collateral for obtaining a loan by the manufacturer. Thus far the Bitcoin "group buy" deals look more like a purchase of a lottery ticket for an office pool that a proper business deal. I'm kinda hoping that this fiasco will give some readers the motivation to familiarize themselves with the commerce and accounting beyond the most basic "consumer" level, or in this case "investomer" level.

Edit: To maybe summarize my post: the choice isn't between "hodl" and "gamble". You always have an intermediate choice of "invest like a pro", which means doing things like "due dilligence", "accounting", "risk analysis" without handholding and safety net from "Fidelity Investments" or other retail brokerage.

[cedivad]

But lying about delivery dates sound like a "prepackaged scam"...How on earth do you promise November delivery when it was clear from the start that they couldn't meet it?

October delivery, please. With the raw wafers being received middle November.

http://hashfast.org/Main_Page#The_Supposedly_Legitimate_Reasons_of_the_First_Delay
http://hashfast.org/List_of_Lies#October_Shipping

I wrote all of that, but I had to read it again yesterday because I basically forgot the majority of it. Time does it. Doublespeak trollers does it.

No "prepackaged bankruptcy scam" would produce a chip like the GN1, which at well over 700GH/s is still by far the fastest BTC ASIC on the planet.

Step 1: Take some sha256 IP, put in millions (not of your money, off course) and make it a reality. Order way more wafers than you actually need, since that wasting money is cool and you need to dominate the market.

Step 2: Package the single die into a 4 die substrate. And don't you dare covering the bare dies with a metal plate or something to help dissipation. Make it the most useless chip ever, by basically putting a kW into a few cm^2. But yes, you can have your favourite trolling arguing that the GN1 is the best chip ever, since that best = fastest. Guess what idiot, I can make you a chip that is 10 times faster than the fastest chip on the market! You only need to dissipiate 10kW from a little more cm^2 this time! AMAZING!

Step 3: If over-storing wafers, over-spending with TSMC and all of that isn't enough, stop shipping orders, raise your payckecks and sleep over the every-minute-more-useless chips, since that you don't give a shit and your best to make them useless wasn't enough.

Step 4: Did I say sleep? Well, sleep. Wait, since that you don't give a shit. You are being paid after all.

Step 5: Oh shit, someone filed for an involuntary (?) bankruptcy, better try to do something to waste even more value. Those useless chips are still worth something after all!

Step 6: Negotiate an impossible deal with an impossible "company", take months to do it, because the chips are better if useless.

Step 7: Take a whole-month-vacation while your lawyers are arguing with the judge that you are not being paid enough money to keep your family afloat (to which the judge answered: "it doesn't compute!"). Hell, if it doesn't compute.


If this isn't a "prepackaged bankruptcy scam", if you really think that, well, I'm sorry, but I can't help you.

[iCEBREAKER]

Step 1: Take some sha256 IP, put in millions (not of your money, off course) and make it a reality. Order way more wafers than you actually need, since that wasting money is cool and you need to dominate the market.

Step 2: Package the single die into a 4 die substrate. And don't you dare covering the bare dies with a metal plate or something to help dissipation. Make it the most useless chip ever, by basically putting a kW into a few cm^2. But yes, you can have your favourite trolling arguing that the GN1 is the best chip ever, since that best = fastest. Guess what idiot, I can make you a chip that is 10 times faster than the fastest chip on the market! You only need to dissipiate 10kW from a little more cm^2 this time! AMAZING!

Step 3: If over-storing wafers, over-spending with TSMC and all of that isn't enough, stop shipping orders, raise your payckecks and sleep over the every-minute-more-useless chips, since that you don't give a shit and your best to make them useless wasn't enough.

Step 4: Did I say sleep? Well, sleep. Wait, since that you don't give a shit. You are being paid after all.

Step 5: Oh shit, someone filed for an involuntary (?) bankruptcy, better try to do something to waste even more value. Those useless chips are still worth something after all!

Step 6: Negotiate an impossible deal with an impossible "company", take months to do it, because the chips are better if useless.

Step 7: Take a whole-month-vacation while your lawyers are arguing with the judge that you are not being paid enough money to keep your family afloat (to which the judge answered: "it doesn't compute!"). Hell, if it doesn't compute.


If this isn't a "prepackaged bankruptcy scam", if you really think that, well, I'm sorry, but I can't help you.

Sure, it's so easy to make a 700+ GH/s BTC ASIC that Superstar Chip Designers at Cointerra could barely match 400GH/s.   

It makes you appear foolish when, in your butthurt rantings, you dismiss the technology in the GN chip/protocol.  There are at least 2 provisional patents...

Sandgate, Chad, and Uniquify are the best in the business.  Cedivad is a nobody who will never design his own ASIC, but loves to put down others' accomplishments.

If this was a "scam" there would be some evidence of that, and the DA would consider filing charges based on the evidence.  But that never happened.

This simply a run of the mill high-tech bankruptcy.  They happen all the time in Silicon Valley.  Many firms must fail so others may succeed.  Creative destruction is part of Capitalism 101.  Go back to Investdurr University if you don't understand that.

It's trolling to keep calling a nearly successful company a scam, when there is zero evidence of criminal intent or actions.

This is, in reality, a bankruptcy, and it would be more productive to accept that and deal with it as a bankruptcy instead of comforting yourself by reciting the litany of Hashfast sins.  Making poor business decisions which doom the company despite its great technology is a boring movie we've all seen before, not a nefarious conspiracy.

Let's repeat: Insolvency and bankruptcy are normal business occurrences.  It is not illegal, nor immoral, nor uncommon, for a firm to experience such conditions. 

You haters are not going to succeed in criminalizing normal business functions.  If risk taking were punished as a form of scam, Silicon Valley would have never existed.  If there were evidence of a "prepackaged bankruptcy scam" the creditors would have shown it to the judge long ago.

[cedivad]

It's trolling to keep calling a nearly successful company a scam, when there is zero evidence of criminal intent or actions.

Two answers for this line only, dear resident troller:

1) If you know HF, a company that didn't create a single dollar worth of value but rather destroyed it all the way along, to be "successful", you must have a really different definition of "successful company", where successful includes scamming, and scamming includes hiding value from creditors. I can agree with you on that.
2) About the "zero evidence", let me quote your friends over at ATM:

I genuinely see ActM as a failed business and not a scam.

Sure, it's so easy to make a 700+ GH/s BTC ASIC that Superstar Chip Designers at Cointerra could barely match 400GH/s.  

1) How do you know that CT's chip is only running at 400GH/s? I was under the impression that they said on their blog since the beginning that the asic met the specifics?
2) How much did CT spent for it's tapeout, in %, vs HF's?
3) How long from when CT had chips in hand to when they delivered? 3 months? "BTC ASIC that Superstar Chip Designers" did it right. "Sandgate, Chad, and Uniquify"... Depends on your definition of right.

It has been a while, so I played your little game and I answered. Let's see what BS you throw out of your keyboard this time.

[iCEBREAKER]

It's trolling to keep calling a nearly successful company a scam, when there is zero evidence of criminal intent or actions.

Two answers for this line only, dear resident troller:

1) If you know HF, a company that didn't create a single dollar worth of value but rather destroyed it all the way along, to be a "successful", you must have a really different definition of "successful company", where successful includes scamming, and scamming includes hiding value from creditors. I can agree with you on that.
2) About the "zero evidence", let me quote your friends over at ATM:

I genuinely see ActM as a failed business and not a scam.

1) You accidentally forgot to include the modifier "nearly" in front of "successful."  HF's technology rocks, but the BTC price/difficulty spike and a couple of missteps doomed the company by draining their coffers.  The lawyers did the rest, by scaring off potential investment and frittering the remains away in court.  No scam, just plain old bad luck and unrecoverable errors.

2) ActM never made a chip, much less the fastest one in the world.  As Chairman Emeritus of the ActM board, I warned them not to try and compete with HF and CT.  They didn't listen.  We are going to find out if ActM was a scam via the results of the MO Security regulator's investigation.  HF never sold securities, and are not under investigation.  zumzero is a nutter, and I have no idea why you'd bring his lunacy into a potentially rational discussion.

   Speaking of "zero evidence"   

"Guilty Until Proven Innocent"
https://www.youtube.com/watch?v=uwjMbs39Nm8

I thought this was America people!
Uhh, yeah, guilty until proven innocent huh?
That's how we workin huh? Okay
...
Uh-huh, okay, I see how we playin
...
Tell me, what you, want from me
Y'all, cats, can't touch me

[Syke]

Sure, it's so easy to make a 700+ GH/s BTC ASIC that Superstar Chip Designers at Cointerra could barely match 400GH/s.   

Yes, it is easy. SHA256 is embarrassingly parallel. Making a big fat chip is just copy/paste. It's just a bad idea due to heat density and other issues. That's why companies that did big chips (HF, CT, BFL) had the biggest problems actually building working miners.

[iCEBREAKER]

Sure, it's so easy to make a 700+ GH/s BTC ASIC that Superstar Chip Designers at Cointerra could barely match 400GH/s.   

Yes, it is easy. SHA256 is embarrassingly parallel. Making a big fat chip is just copy/paste. It's just a bad idea due to heat density and other issues. That's why companies that did big chips (HF, CT, BFL) had the biggest problems actually building working miners.

We all know BTC mining is trivially parallelizable in theory; nobody is impressed with you dragging out that old chestnut.

What you seem unaware of is the 'more is different' principle, although you came close to finding it with your passing reference to "other issues."

http://en.wikipedia.org/wiki/Philip_Warren_Anderson

Your most embarrassing omission is leaving KnC off the list of "companies that did big chips."

Of course you had to pointedly ignore KnC, because as the most successful BTC ASIC company ever, their example proves your latter claim demonstrably false.

[jimmothy]

Of course you had to pointedly ignore KnC, because as the most successful BTC ASIC company ever, their example proves your latter claim demonstrably false.

You forgot to add "/s" so people don't think you're being serious.

[iCEBREAKER]

Of course you had to pointedly ignore KnC, because as the most successful BTC ASIC company ever, their example proves your latter claim demonstrably false.

You forgot to add "/s" so people don't think you're being serious.

So which BTC ASIC company is more successful than KnC?

And if KnC is just another big hot chip failure like HF/CT/BFL, why did Syke accidentally forget to include them in his list?

[notbatman]

I look at my situation with Black Arrow and it looks bleak.

Then I come to this thread and see that there are those that don't even get a fancy space heater with bells & whistles.

The X3 can play flappy bird  .

[jimmothy]

Of course you had to pointedly ignore KnC, because as the most successful BTC ASIC company ever, their example proves your latter claim demonstrably false.

You forgot to add "/s" so people don't think you're being serious.

So which BTC ASIC company is more successful than KnC?

Bitfury, Bitmain, Avalon, Asicminer, Lketc..

Nobody wants to buy a giant chip that requires $150 worth of cooling when you could use smaller chips with ~$30 worth of cooling.

[pmorici]

So which BTC ASIC company is more successful than KnC?

Bitmain.  They are the only ones that have consistently met their deadlines and specs and in the one case where they had a small miss on specs they compensated customers proportionately.  Spondoolies is the only other manufacture even close.  It's notable that both of these companies appear to have had their R&D funded by private investment.

All the other manufactures in hindsight appear to be thinly veiled schemes to use customer money to finance R&D which they then use to build private mines once all the risk has been eliminated with other people's money.

Hashfast lead everyone to believe that their R&D had been privately funded but we now know that was false and they were using customer money to fund and operate the company.  Money which was not rightfully theirs since they hadn't yet shipped any products.

http://en.wikipedia.org/wiki/Revenue_recognition

[iCEBREAKER]

Of course you had to pointedly ignore KnC, because as the most successful BTC ASIC company ever, their example proves your latter claim demonstrably false.

You forgot to add "/s" so people don't think you're being serious.

So which BTC ASIC company is more successful than KnC?

Bitfury, Bitmain, Avalon, Asicminer, Lketc..

Nobody wants to buy a giant chip that requires $150 worth of cooling when you could use smaller chips with ~$30 worth of cooling.

Bitfury...maybe.  KnC already has like $20 million in the bank, while BitFury is borrowing $20 million in venture capital.

Let's not forget BitFury also had "problems actually building working miners" even with its tiny chip.  They had to run it at half speed, remember?

The rest?  Not a chance.  Avalon pissed off at least as many customers as HF.  Bitmain is still a baby.  And AM's latest chip is a dud.

When you have to pay for electricity and rack space, the trivial increased cooling cost is overwhelmed by savings accrued by much higher density.

Regardless, leaving KnC off your list that purported to demonstrate the non-viability of big hot chips was a silly mistake to make.   

[jimmothy]

So which BTC ASIC company is more successful than KnC?

Bitfury, Bitmain, Avalon, Asicminer, Lketc..

Nobody wants to buy a giant chip that requires $150 worth of cooling when you could use smaller chips with ~$30 worth of cooling.

Bitfury...maybe.  KnC already has like $20 million in the bank, while BitFury is borrowing $20 million in venture capital.

Let's not forget BitFury also had "problems actually building working miners" even with its tiny chip.  They had to run it at half speed, remember?

Do you really think Bitfury has no money in the bank? They raised $20m on a $270m total valuation. The only reason KNC/HF didn't raise $20m is because nobody wants to invest in them.

Even with a chip that never achieved the full 5GH they still managed to sell/deploy enough hardware to account for 40% of the network.

The rest?  Not a chance.  Avalon pissed off at least as many customers as HF.  Bitmain is still a baby.  And AM's latest chip is a dud.

Avalon still managed to sell 9PH (not including self mining) worth of overpriced hardware after "pissing off as many customers as HF" and their first gen was hugely successful.

AM already turned a profit of 210,000 btc and their latest chip is more energy/cost efficient than HF/KNC's chips.

Bitmain has managed to maintain 20% of the network for around half a year now. If Bitmain is a baby does that make KNC a fetus?

I don't know how you can say KNC is the most successful company while they have an incredible amount of pissed off customers demanding refunds and yield issues so bad they can't even ship to the few who still wanted their neptunes.

When you have to pay for electricity and rack space, the trivial increased cooling cost is overwhelmed by savings accrued by much higher density.

The savings on cooling are not nearly trivial when they make up ~25% of your production costs. Also the most dense miner to date (sp30) uses cheap extruded heatsinks.

Regardless, leaving KnC off your list that purported to demonstrate the non-viability of big hot chips was a silly mistake to make.

I agree he should have included KNC in the list of failures.

[Syke]

Regardless, leaving KnC off your list that purported to demonstrate the non-viability of big hot chips was a silly mistake to make.   

No one said big chips are impossible. Just that they are fraught with difficulties, as proven by the fact that most big chips have had huge problems making working miners, while most small chips built working miners quickly and easily.