Black Arrow 28nm 100Ghash Bitcoin ASIC from $0.49/GH/s

[jimmothy]

Thanks david. Do we have the full minion specs yet? 0.92w/gh is not great but its not bad either. Can the chips reach better efficiency by underclocking?

[sorehammer]

Cheers David  that's good. 😊

[y_boonstra]

BA.

im very sorry that i have to be the negative person here again, but WTF happen your chip is at the same point as was your first design back in feb.. these specs are terrible !!
your in the same range as an ant miner, and yet your price is higher + you don't have the units in stock.

surly you can't deliver the machines with these kind of specs then your screwing your costumers over, are you planning to replace these chips with new ones if the design is reviewd, iff not your screwing batch 1 and batch 2 ppl over big time.

at these numbers the machine will be worthless on diff 30 bil... this can't be true

your machine is running at 1,5W/Gh at wall while your were telling us it WILL preform around 0,75W/Gh on chip level @ T/T corner (120Mhz) !!!!

YOUR CUTTING OUR POTENTIAL MINING TIME IN HALF WITH THESE SPECS..

it almost sounds like your rushing out the machines before PPL realize its crap and start asking refunds

[jimmothy]

150w @ 120gh/s is 1.25w/gh at the wall (compared to 1w/gh had they hit spec)

Anyways what do you want them to do? Send the chips back to design stage and let us wait another 3 months? I'd rather just get my asic now.

[y_boonstra]

150w @ 120gh/s is 1.25w/gh at the wall (compared to 1w/gh had they hit spec)

Anyways what do you want them to do? Send the chips back to design stage and let us wait another 3 months? I'd rather just get my asic now.

it's almost the same W/Gh ratio as an ant miner wich is cheaper and available now.

[Syke]

150w @ 120gh/s is 1.25w/gh at the wall (compared to 1w/gh had they hit spec)

Anyways what do you want them to do? Send the chips back to design stage and let us wait another 3 months? I'd rather just get my asic now.

it's almost the same W/Gh ratio as an ant miner wich is cheaper and available now.

It's worse than both Bitmain and Bitfury, 55nm chips from last year!

[coolmusic]

150w @ 120gh/s is 1.25w/gh at the wall (compared to 1w/gh had they hit spec)

Anyways what do you want them to do? Send the chips back to design stage and let us wait another 3 months? I'd rather just get my asic now.

it's almost the same W/Gh ratio as an ant miner wich is cheaper and available now.

Bitmain is 2W/Ghs in NORMAL mode
Blackarrow is 1.25w/Ghs OVERCLOCKED

[coolmusic]

They mean the Antminer S2, not the S1. The S2 is basically the same thing as one of the X-3 subunits, but has been available for a bit already.

thats true but my point is wait for the final numbers to make comparisions!

[jimmothy]

150w @ 120gh/s is 1.25w/gh at the wall (compared to 1w/gh had they hit spec)

Anyways what do you want them to do? Send the chips back to design stage and let us wait another 3 months? I'd rather just get my asic now.

it's almost the same W/Gh ratio as an ant miner wich is cheaper and available now.

Bitmain is 2W/Ghs in NORMAL mode
Blackarrow is 1.25w/Ghs OVERCLOCKED

They mean the Antminer S2, not the S1. The S2 is basically the same thing as one of the X-3 subunits, but has been available for a bit already.

S2 is $2500 for 1th or $2.5/gh

BA is $6000 for 3.6th or $1.7/gh

[coolmusic]

They're shiping in less than a week i wouldn't expect a large change in numbers.
[/quote]

ok! so, what is the minion power comsumption when run at 100 Ghs/s?

[Vcore]

I think I can clear some of the confusion (I am not related to them but I know how the circuits/components involved work) .

The choke/inductor is not from the power supply but from the 12 to 0.85v dc-dc converter in the hashboard and that is good news as they are not made yet.

They should be able to provide a 10 to 15% improvement in overall efficiency as long as the rest of the dc-dc converter design is really optimal. Designing a dc-dc converter for those powers is really hard and at 200A a tiny internal resistance anywhere is catastrophic (and inductors are a prime candidate for unwanted internal resistance).

Also knowing the real asic specifications they might extract some extra stability/overclockeability by adjusting the vdroop and some extra efficiency by adjusting frequency and gate drive parameters at the dc-dc converter so there is stil a chance for some improvement.

Also you can't make a power supply or dc-dc converter 100% efficient and we are going to have at least two between the wall and the asics. 1st one turns mains into 12V and 2nd one turns 12V into the asic vcc (0.85V if i am right). So lets say that both those converters are 90% efficient wich would we an awesome job, 0.9*0.9=0.81 so roughly 20% more power at the wall than at the asic would be an awesome job. If they have active pfc they might have even another dc-dc converter.

[fivejonnyfive]

You know, I'm frustrated with the delays - frustrated as all hell just like the rest of us.

But honestly? 150W for 120GH is pretty palatable. It's barely above the power consumption of the antminer S2 - and I don't doubt that these numbers can be improved upon.

The compensation package for batch one should be 2x the units. No question about that. I can get a 180gh antminer for less than what I paid for this 100gh x1. Hell - I already did.

But like, the power consumption seems pretty decent and I bet the final numbers will be better

S2 is $2500 for 1th or $2.5/gh

BA is $6000 for 3.6th or $1.7/gh

Where the hell are you getting that 3.6th number from? Didn't they say they've got 120gh a chip?  The x3 would get 2400th in that case, no? $2.5/gh in that case.....

[tempestb]

Thanks VCore that explains a lot.

How many chips are on the board?  Would it halve the inefficiency with each chip added?  We're looking at 150watts at the wall for one chip currently, I'm thinking that won't translate to 300watts with two chips on the board.  What do you think?

[notindexed]

Dear Black Arrow,

I have a solution for your batch 1 X1 preorder folks!

Example here is a 10 units X1s preorder:

Your current compensation:
- Ship 10 units & 5 units w/o PSU (no real statement on how they will handle shipping, no PSU really? and will this delay shipping even more?)
- Offer some kind of cloudhashing using instock X3s to compensate the delay vs. X3 shipping (pain in the butt to setup/distribute evenly, inc over 9k tickets etc..)

A better solution for your backers and yourselves as it saves you time, headaches, resources and rep.:
- Ship 1 X3 instead of the X1s "on time" (less shippingcosts & the above is about the price of an X3 & you can ship earlier)

tl;dr

Ship 1 X3 per 10 units X1 preorder and ofc adjust X3 preordercompensation accordingly.

[Dot99lr]

b. The expected loss on the DC/DC should be around 10%. We have identified We identified that part of the power used is lost by the inductor. A new customised inductor is scheduled to arrive in one week.



Wouldn't that mean a week to change the inductor, thereby delaying shipping by at least a week? They said shipping early next week. Am I missing something?

[jimmothy]

How many X-3 units are you keeping for your personal collection?

Bobsag said that when you announced your data center it would be used for the compensation package, that doesn't need to happen anymore so how many units are you guys sticking in there that you refuse to compensate us with, because that's all you can do?

EDIT: what is your current hashrate from the chips that are in now?

Why does it matter how many x3s they will have in their DC?

Did you think they were going to build exactly as many x3s as needed for their rent-a-minion orders and no more?

Seems to me like your just grasping at straws for something to be mad about.

[jimmothy]

I absolutely did not think they wouldn't keep some, but I think we deserve to know how much they are going to make off of our investment (and our loss)
And honestly at this point you just seem like your trolling (not only this thread, but most of the ones you post in)

EDIT: yea responses on about 90% of your posts point out your trolling, while the other 10% is just ignored.

You didn't think they would make a profit off their business?

I understand you are not happy about the delay but expecting them to go in to bankruptcy so you can get the maximum possible compensation makes no sense.

[ChuckBuck]

I absolutely did not think they wouldn't keep some, but I think we deserve to know how much they are going to make off of our investment (and our loss)
And honestly at this point you just seem like your trolling (not only this thread, but most of the ones you post in)

EDIT: yea responses on about 90% of your posts point out your trolling, while the other 10% is just ignored.

You didn't think they would make a profit off their business?

I understand you are not happy about the delay but expecting them to go in to bankruptcy so you can get the maximum possible compensation makes no sense.

Yea, but it's okay for it's investomers to go into bankruptcy or major debt because a company's gross incompetence.

Makes sense!

[tempestb]

They aren't mining anything in their data center because they don't have a complete board yet, it won't be done until tomorrow for testing.  If you read the update, the manufacturer of the circuit board hadn't started it on time. 

I dunno why you don't just ask for a refund if you're so upset that you don't have your miner yet.

[bobsag3]

I absolutely did not think they wouldn't keep some, but I think we deserve to know how much they are going to make off of our investment (and our loss)
And honestly at this point you just seem like your trolling (not only this thread, but most of the ones you post in)

EDIT: yea responses on about 90% of your posts point out your trolling, while the other 10% is just ignored.

You didn't think they would make a profit off their business?

I understand you are not happy about the delay but expecting them to go in to bankruptcy so you can get the maximum possible compensation makes no sense.

Oops, was watching baseball/draft and just realized that was phrased badly.  Of course I expect them to keep some and make a profit, but this first batch was entirely funded by our orders, and they have said numerous times that there isn't anything else they can do.  And yet, they are keeping untold numbers of chips, currently mining with untold number of chips, and delaying things further and further, without giving clear roads.

There is still not a timeline on when the X-1's will start to ship, and on top of that, the LCD (which apparently was such a nice feature, even though 99% of people will set up the miner either some place they visit rarely, or will never look at it) will not even work with what is looking to be the entire first batch.  Add on that they are missing specs (doesn't matter if better than antminer, still on par with spondoolies, and worse than AsicMiner (proven) and BFL (allegedly)), and I think we are entitled to some answers/compensation.  The only reason they will make a profit is because we paid for their development and run of chips.

You cant tell me you enjoy being lied to, strung along, lost money, and finally screwed over with the compensation?  And anyone who says they do not have to compensate anyone, you are plain wrong.  They clearly said at the beginning that they would match any price per GH with extra units shipped, now it seems that doesn't really matter.  Or perhaps BA knows that since they haven't shipped, it isn't worth talking about anyway.

Also, the entire delay was due to chip optimization, and yet, we are higher than the specs that were released before their "optimization process".

Bobsag, I hope you demand to see every inch of their business and let us know how many chips are currently "being tested" at their facility.

Im not leaving until I can take away a finished X3 [1 box at least, bringing an empty suitcase], and if I can a X1 without a case. Will also take as many pictures as I can upload over vpn.

I have not left yet, waiting for final go from them to leave.