Topic: BitForce SC - full custom ASIC - page 15. (Read 52532 times)

It's funny you should mention Kickstarter. One of the big Kickstarter projects took in a cool third of a million dollars and still hasn't delivered anything. Its backers have been feeling quite thoroughly scammed for a while:

http://www.kickstarter.com/projects/zioneyez/eyeztm-by-zioneyez-hd-video-recording-glasses-for/comments

I didn't quote the rest of your post because this is really the heart of the matter. 

This argument is pretty much the only valid argument I've seen come out of anyone (Yes, Mem included) and has never really been a point of contention.  Yes, they have failed their shipping estimates.  They could have and should have handled those far better... but other than failing to meet shipping estimates, I have yet to see any convincing argument that BFL has dealt with customers in anything less than a respectable manner.

Again, the initial estimates for the Single were flawed, and that came from a healthy dose of ignorance in so far as what they were dealing with and not any intent to deceive or defraud.  You can also include that if you want, but I personally do not count that as a flaw unless the practice continue(s/d) with the Mini-Rig and the SC.  As a first run and really the first "mass produced" FPGA solution, a miss-step like that is fairly minor in the grand scheme of things, especially in light of the fact that they were happy to refund pre-orders for customers who were dissatisfied.

For something they could initially easily sell at over 10x or 100x the cost to produce, I imagine they can find solutions for that.
Hint:

Again, Paypal, like MC/Visa, does not make nor enforce laws.  So citing Paypal policy in no way backs up your claim that it's somehow "illegal."  So again, your credibility is non-existent until you can prove your claims.  You have posted nothing "I don't want to hear," you just posted "nothing," other than citing policies from companies that have no ability to make or enforce the laws you are claiming exist.

Are they shipping 50 units of BFL singles a day?  It seems like they are around 10 units a day shipping average.  Why would these new units be able to be shipped 5X faster? These have to have some human interaction to build which slows down the building.

At 10 units per day shipped @5GH each would add 50GH to the network per day, 250GH per week, 1TH per month.  It will take 1 year from the date they start shipping until the network hashrate doubles.  I think getting 10 units shipped a day would be tough to keep up.  You run out of cases, or power supplies, or cables and orders get backed up. 

Nice to see you acknowledge the issue at last, and  Im really looking forward to finding out how you intend to pull that off, because I cant see way that wouldnt undermine your own profitability significantly as long as you maintain a sole supplier status, and I cant see a way at all if/when an asic competitor emerges. in the former case I can only see artificially limiting your own sales as a solution, in the latter case youd have to compensate your customers for your competitors sales.. not very likely..

Given that we're now at 11TH/s it's a five fold increase in difficulty which, compound with block reward halving at year end, becomes a ten fold increase.

If BTC reaches 50 USD then all will be ok, otherwise...

spiccioli

I really hope so.

Let's wait until the announcement before we continuing this discussion.

Can you tell us if the announcement is early or late June.

Our general (and strict) policy is to make sure our clients recover their investment and make proper profit in a timely manner.
I cannot explicitly disclose more data prior to our announcement, but it would be sufficient to say that whatever path we choose
must and will be a 'win-win' path (even if it translates into loss of profit for us). Any other strategy will result in loss-of-client
which is certainly not what we're after.


Regards,
BF Labs Inc.

So my best guess is that first batch of this "wondies" will be 5GH/s@10W per chip. Propably at 10k volume. That means 50TH extra hash power. 50 units sold per day. 200 days....

yoyo implies 2 directions. I agree, that wont happen. But if you are implying you will hold prices steady even as difficulty skyrockets, I think potential buyers would like to hear that said explicitly, because I dont believe it for a second.


Yes, but that offer will hardly cost you anything. Mind you, I think its a good idea, and Ill credit you for it,  but its nothing like the tradeoff you will have to make once difficulty begins skyrocketing.

A company cannot simply play Yo-Yo with its product pricing. This will result into loss of respect from the clients, and gives
company a bad name. Our history clearly shows that we have not done such thing with our singles or Mini-Rigs. Even now that
we are announcing ASIC, we have the trade-in program to make sure our clients do not suffer investment loss to a newer
generation of mining platform.


Regards,
BF Labs Inc.

No, it will not be the same.  CPUs may have a low variable cost just like your asic, but the market price of CPUs was never determined by the bitcoin difficulty. Intel wasnt going to lower prices because bitcoin miners where no longer buying, nor could AMD charge 100x their production cost because bitcoin mining made it profitable. It was the opposite, the price of cpus (and later gpus and fpgas) was and still is set by factors unrelated to bitcoin, so that price/MH resulted in a (more or less predictable) bitcoin difficulty for a given exchange rate.

Like GPUs and FGPAs before them, given the inherent performance and performance/W advantages, ASIC prices will soon determine future difficulty. But unlike cpus, gpus or fgpas, future difficulty will also determine future ASIC pricing. That makes it 100% different.

Does the ASIC announcement officially include ASIC-powered Single 2.0 units?

What ASIC does is that it will increase the average processing power of the miners in the network. In CPU mining era, average processing
power of an average miner would've been for example 40MH/s. Those who invested more, could've gone as high as 200MH/s with several
motherboards and CPUs (or as high as their investment allowed them). This relativity was maintained when network switched to GPU.
Today average processing power of an average miner could be 1GH/s, and miners who invested more could be at 5-8 GH/s  (average
well-funded miner). The same will happen when ASICs hit the market...


Regards,
BF Labs Inc.

I think it would probably be better for miners if the high-end mining gear is limited to FPGAs, since the cost of the chips are kept high.

When I purchased my BFL Singles, I thought ASICs would still be several years away.  I guess I was wrong.  If I knew at the time ASICs were less than a year away, I might not have bought the FPGAs.

But I guess you can't put the genie back in the bottle.  I'll likely be in for some ASICs, so I can be part of the game, but the risk involved just increased--a lot.  And the total dollar amount I'm willing to invest in mining gear just decreased accordingly.

Regardless of the cost of ASIC mining equipment, I think the pool of miners is still going to be largely limited to bitcoin nerds and geeks, so with the customer base limited, a lot of sales will need to be repeat business.  It's in BFL's best interest to balance things out in such a way as to not alienate their customers.  Key word being balance.

And how could you? And not only that, you would need to know how many will be sold in the  future. Thats what I said all along, without BFL telling you everything they wouldnt want you to know, you can not make an informed decision and buying their product at *any* price is purely a gamble; one that will almost certainly lose you money if BFL pursues maximum revenue. Its quite the opposite of BFL and miners interests  being perfectly aligned, they are diametrically opposed.


And what makes you think if they announce a 1 GH unit next month, that they will not up that to 5 or 50GH later on? I mean, they could just lower the price as difficulty increases, but if they are in it for the money, the smart thing to do is fuse off miners in those chips and start by selling truckloads of Asics that are roughly comparable to singles in performance/$ (but far superior in performance/W) and steadily increase the hashrate of these units,  instead of lowering the price. It amounts to the same thing, both lower performance/$ but its actually easier to pull off considering the non asic costs of every unit. Oh and then they could even sell you an "upgrade", because they care so much about their customers.

It may seem weird if you didnt think about it, but if the chips cost on the order of $10 to produce, it makes perfect sense to throttle them initially and slowly unleash their true potential. Then you make your estimates to make your "informed" purchase, and you will get burned terribly.

All together now 

And perhaps more pop-ular!

(Man, that was horrible. I'd best be getting to sleep.)

I wonder if I should trade DMC shares for popcorn. Might be more cost effective.