HashFast announces specs for new ASIC: 400GH/s

[dopamine]

"Our updated estimate is that the rigs will begin shipping in mid-November"

Now that we wont get RIO what will hashfast do to compensate for our lost?

1.refund
2. Add more hashrate to MMP
3. Do nothing
4. Ship with MMP

[Visesrion]

"Our updated estimate is that the rigs will begin shipping in mid-November"

Now that we wont get RIO what will hashfast do to compensate for our lost?

1.refund
2. Add more hashrate to MMP
3. Do nothing
4. Ship with MMP

How about this :

Well, the main thing here is when are going to ship the 2 bach sierras and upgrades.
Why dont HF goes straigh with just sierras and leave the babyjets ? It seems they just focus in sierras design indeed.
The boards and chips are the same, so why not ship this first customers a sierra ?

As I suppose they are going to ship this 2 batch units really close to the first batch, so it would be a way to reward this first customers for the delay, to reward also for the sales they made where you can buy x3 hashing power at the same price and they may earn a lot in long term if they dont have to send MPP chips/boards by sending more power.

[Gyrsur]

KNC had a delay, did not inform about or to late.

FYI, KNC was supposed to ship in September, but the first customers received them in early October. I received my first unit Oct 4th. Delays are delays, no disagreement, but KNC had a fairly minimal delay.

but they had a long, long time (August, September) without nearly nothing substantial comments about the progress regarding their ASIC chips and that was not very satisfying.

[Hoofprint]

What are the chances of ROI with this delay and the MPP kicking in 3 months later, wont batch 2 and 3 sierra's be online by that point ?  

[ninjarobot]

The majority of the project is on track and, in many cases, ahead of schedule. However, one important component, the substrates, is going to take a few weeks longer than we wanted.

A few weeks delay matters *a lot* when your mining income looks like an asymptotic curve:



Since the MPP timer only kicks in when the units are shipped any delay here will propagate to the additional units. As such it is unlikely that even with full MPP protection customers will ever break even.

Also note that Batch 2 (late Nov) and 3 (Dec) will ship very shortly after Batch 1, putting further pressure on the 'profitability' of Batch 1 customers.

[cedivad]

2 to 3 difficulty adjustments later = 2x to 4x the difficulty. Do i have to say how useless and overpriced these units are at that time? Batch 1 customers, MPP or not, will lose half or more of their investment. Probably more.

Is HashFast gonna do something over this or what?

[joshv06]

2 to 3 difficulty adjustments later = 2x to 4x the difficulty. Do i have to say how useless and overpriced these units are at that time? Batch 1 customers, MPP or not, will lose half or more of their investment. Probably more.

Is HashFast gonna do something over this or what?

Yeah, who's picking up the tab? We are basically paying double. November Batch rates. On top of that, when MPP hits 3 months from now, difficulty will be sky high, and we would have to pay 4x electricity costs.

My solution:

Send all batch 1 people 2x-3x the power we ordered in Sierras.

[cedivad]

2 to 3 difficulty adjustments later = 2x to 4x the difficulty. Do i have to say how useless and overpriced these units are at that time? Batch 1 customers, MPP or not, will lose half or more of their investment. Probably more.

Is HashFast gonna do something over this or what?

Yeah, who's picking up the tab? We are basically paying double. November Batch rates.

Anyway, if it's delayed up to (say) 3 years from now, when the hashrate is stable, our return would be sure (thanks to MPP). The only case in where the mpp works, is that one.

[Visesrion]

My solution:

Send all batch 1 people 2x-3x the power we ordered in Sierras.

That is what I suggested some pages ago 

I have sent them this suggestion to them. I will come back when they answer me

[itod]

The only way that HF can look less bad after today's announcement is the statement that MPP begins November 1st.

That's a minimum they can do.

[Paladin69]

The only way that HF can look less bad after today's announcement is the statement that MPP begins November 1st.

That's a minimum they can do.

I feel bad for second batch Sierra's.  End of January is light-years away in difficulty changes.

That guy who bought 200 Baby Faps is really screwed on ROI.  He spent a boat load of coin for that.

[dopamine]

WE WANT ANSWERS you have our bitcoins!

[cedivad]

WE WANT ANSWERS you have our bitcoins!

4+ millions of dollars of them.

[WastedLTC]

WE WANT ANSWERS you have our bitcoins!

4+ millions of dollars of them.

expensive way to launder 4 million..   

[cedivad]

I'm contacting a lawyer in CA to see what can be done. PM me if you are interested in sharing the fees.
(it's a preorder, and i'm demanding for a refund)

[Threader]

HashFast is looking HashKindaFast but hope it doesn't turn into HashSlow  

[pwnyboy]

I don't think enough attention has been paid to the actual update given.  And I think Hashfast customers might be letting them off WAY too easily.  What are these "substrates"?  Are we talking about the silicon substrate of the ASICs themselves?  If yes, their statement would make a lot more sense:

Once the substrates arrive, we can place the silicon and finish production.

If I understand correctly, the substrate is lost in space, there may be some ETA for it, but they're not telling you what it is.  And whenever they magically receive this substrate, they're going to finish packaging the ASICs and just assume everything went according to plan.  I mean specifically that the PCBs and everything else works the way they're supposed to, which they can't possibly know until they have an ASIC in hand.  Do I have that right?

So set sail for more fail.  At this point they were supposed to have ASICs in-hand, boards completely tested, and starting to ship.  If/when additional problems are encountered with parts external to the ASIC, you can expect more delays.  A quick recommendation to Hashfast: perhaps when you have engineering-related issues, rather than sending out a talking head, you should instead send your engineer to update your customers and address their concerns.

[Puppet]

WE WANT ANSWERS you have our bitcoins!

4+ millions of dollars of them.

Out of curiousity, where did you get that number from?
For the record, I suspect its *far* more than that. So far KnC seems on track to ship almost 3000 "boxes" for the "september" orders alone (https://www.kncminer.com/news I counted 2385 boxed and one 'good day' of production to go), that would be well over $10M for just one month, and KnC always seemed like the smaller underdog to me.

[Puppet]

If I understand correctly, the substrate is lost in space, there may be some ETA for it, but they're not telling you what it is.  

Its a rather crucial part of the chip package. Its like a mini PCB.

And whenever they magically receive this substrate, they're going to finish packaging the ASICs and just assume everything went according to plan.  I mean specifically that the PCBs and everything else works the way they're supposed to, which they can't possibly know until they have an ASIC in hand.  Do I have that right?

Yep. Though to be fair, thats exactly how it went with KnC, and  I assume they did plenty of testing with an FPGA. Of course its entirely possible (likely even) they will run in to more hickups, like the firmware issues KnC had. If they dont get packaged chips in the next 10 days, I seriously doubt they will ship anything before the tail end of November.

[DeathAndTaxes]

What are these "substrates"?  Are we talking about the silicon substrate of the ASICs themselves?  If yes, their statement would make a lot more sense

Just a random googled chip image not HashFast processor.

The shinny metal part is the actual die (produced on wafers at a fab like TMSC, cut into individual dies).
The greenish particle board looking part is the substrate.

The entire thing is what most people call "the chip".  For high powered processors the die and substrate will be covered by a metal lid which acts as a heat spreader.

So whatever the issue with the substrate is, they are needed in the step called "packaging" where completed wafers are cut into dies, attached to substrate to form complete packages ("chips").

Processor/chip = silicon die + substrate + package lid + packaging labor