IM just reporting my experience; I have been using bitcurex for about 2 years now and never had any major issues. Slow responses, yes, but this time they were quite fast.

BTC deposits are clearly automated, since they appear within seconds of getting 6 confirmations. If your deposit doesnt show up, I can only guess you messed up somehow. Care to post the tx id?

I appreciate the feedback. Ill let this run for a while longer and then make some adjustments based on what I read. But so far, it seems we are all broadly in agreement, which is a good thing .

No. Did I say you said it? No.
Please read carefully.


Sigh.

I never assumed 100s of PH wil be coming online in 6 months, please look up the definition of a hypothetical.
Moreover, the hypothetical didnt state that 100's of PH would be online in 6 months, merely that during the next 6 months we would see them being sold and being deployed. Note the present continuous.
Finally I am saying the hypothetical is at least plausible, and more plausible than the other hypothetical I gave of only a few dozen PH. Considering you agree at the current rate without anyone known to be shipping 28nm gear, we could already be on track to add 50PH its hard to see how you are arguing against the plausibility of that number multiplying considering about 10 vendors are expected to start shipping latest gen miners in the next months and currently there are none (known).

Cartels wont work for bitcoin miners. Mining is a zero sum game, so asic vendors are competing with themselves even if there was no one else.  So even a monopolist can not continue to charge prices that will make their product appear unable to ever ROI. Every TH they add to the network reduces the market value of their unsold products.

Its a hypothetical, not a prediction. 12 PH seems even more absurd. FWIW we seem to be adding ~3-400 TH daily at this point, before anyone we know is shipping. If that were to continue you are already looking at 100PH per year. You dont think the pace will pick up once HF, CT, Bitmine, BFL etc will start shipping?

Looks similar, but that fan only has 2 "flat" sides, the one in the rendering has 4. That said, its just a rendering and you can probably buy similar fans in any size you want.

Im guessing that is a 140mm fan, if for no other reason than Ive never seen a fan of that shape in other sizes.
Again, design looks quite sleek and sensible. Its nice of BA to use as big (and therefore silent) fan as possible, even though Im sure it would cost more than a noisy, higher rpm 80mm fan.

Got a response to my ticket today with a promise to credit the amount. Yes, it was in broken english, and yes it would have been better if they had done it automatically, but I still got a satisfactory response within 2 working days, which in bitcoin world is actually spectacularly good.  Maybe they did increase support staff after all.

Thats an excellent point I had overseen/forgotten. Indeed, their hosting service was quite a disaster (and way overpriced). Whats the situation on that now? Are they at all hosting for customers ? Did they refund the hosting and allow you to self host your miner?

Depends how you define better. Bitmain isnt taking preorder "investments" and is instead selling from stock, that has to be an advantage, since there is basically no risk of long delays or missing specs. Of course what they deliver is less impressive than what KnC has done and might do, but the point of this thread is not to assess how good those vendors offers are in terms of price or efficiency or whatever, just how likely they are to deliver on their promises. In that sense, its going to be difficult to do better than Bitmain.

While competition is a factor, you cant talk about prices without talking about future difficulty. And future difficulty is a direct result of sales volume. That is the big unknown (aside from BTC price obviously). If all asic vendors combined only sold a few dozen PH, then Im sure there is room for prices to go up, at least for some time. If otoh, over the next 6 months we will see 100's of PH are being shipped and deployed, and BTC price remains roughly where it is now,  I dont see much of a market for $3/GH miners anymore.

LCD seems to have disappeared, or is that BA logo plate really a display ? (and then probably also on their X3's ?)
If so, that would be pretty cool.

I thought this might be useful for newbies, to give them an idea of the track record / reputation of the various bitcoin asic vendors.

Consider this work in progress, its based on things I read, I dont have first hand experience with any of them, let alone all of them,  but a single person's experience wouldnt be very representative anyway. Im very much open to changing/updating this based on feedback posted in this thread, particularly for vendors Im not too familiar with, but I will weigh your opinion/experience against your own reputation. Ie, opinions and stated experience of newbie posters or known trolls wont carry the same weight as those of trusted long time members.

This is not meant to be a scientific study, its my personal and therefore subjective interpretation of what I know about these companies. I dont have a bone in this fight, I hold no shares nor preorders for any of them, which should make me unbiased, but Im a human being with my own opinions so YMMV.

In alphabetical order:

AMT 2/5
Doesnt produce asics, but assembles miners based on bitfury and bitmine chips. Have shipped a handful (possibly as few as 2) Bitfury/technobit based 55nm miners so far. Currently 6+ weeks overdue on their bitfury orders and advertising highly implausible specs and delivery dates for their Bitmine based 28nm gear. Customer support is very poor, reportedly very rude and mostly non responsive. Have a very unfortunate (published) policy of giving 5% discounts and shipping priority to shill accounts posting in their favor, making any posted positive customer experience automatically suspect, particularly from new (often fake) accounts.

Avalon ?/5
Delivered on their promises with batch 1 of their first gen (110nm) devices, royally screwed their customer on batch 2. Their second generation 55nm asic appears to be shipping from stock, but is aimed more at system integrators than end users. Ive not heard enough recent feedback to make a judgement, but it appears they may be trying to salvage what little is left of their reputation. Specs for their 55nm chip suggest their design is less power efficient than current competitors, but it should also be pointed out their miners are quite price competitive atm.

Asicminer
Asicminer is a startup that was publicly funded with bitcoins, through the issuing of btc denominated shares that can be traded, similar to ActM/VMC, but with a proven trackrecord. Asicminer was a pioneer in bitcoin asics, along with Avalon. They mostly self mine and at one time they owned about 1/3 of the network. Currently they only have an almost obsolete chip. A new generation is supposed to be under development, but AFAIK, they are no longer selling directly to end users atm. Their products like the blockerupter and cube are still available from some third parties, as well as various boards based on their chip.  Keep in mind these miners have very high power consumption for their hashrate and especially the USB sticks at todays difficulty are all but obsolete.

update: asicminer announced tapeout of their 45nm follow up, due in May. It has pretty incredible low power usage specs, lower than any 28nm chip.

BFL 1.5/5
Have delivered several PH of 65nm hardware, but are notorious for delivering extremely late and often underspec. Customer support is completely inadequate and incapable of handling the volume they receive . Their own forum contains countless reports of DOA's and machines breaking down after only a few weeks. In particular, dying/sparking/burning PSU's. TO be fair, BFL is one of the largest suppliers out there, its difficult to estimate what % of their sales these stories represent. Their 28nm product is already 2-3 months late and no silicon in sight so far.

Bitfury -/5
Have shipped several PH worth of 55nm asics, delivered pretty much on time but below spec. I have not read anything on customer service, so I will refrain from assessing that. Currently seems to focus on working with third party providers like Technobit so may not be relevant anyway. Performance and power efficiency of their 55nm product was revolutionary at launch and still is impressive today, rivaling that of many 28nm designs.

Bitmain 5/5
Relatively small and new vendor, doesnt do preorders, instead sells mostly through auctions,  shipping quickly and from stock, specs are as promised, few reports of DOA or needs to RMA. Bitmain doesnt even have a website yet (sales happen through auctions on this forum), but customer support appears very good from what little Ive read about it. Like Bitfury, Bitmain "only" has a 55nm chip to date, but its also competitive in power efficiency with Bitfury and current 28nm alternatives.

Bitmine.ch 3/5
Havent shipped any 28nm miners yet, currently a few weeks overdue on their initial promises. Have shown working silicon and might be close to shipping. Too early to assess. (note: I believe bitmine used to ship avalon based gear? If so, some feedback would be welcome, Im not aware of their track record).

update: although more than a month late, full scale production is underway and shipping supposedly imminent.

Black Arrow 3/5
Have only shipped FPGA miners and recently Bitfury based gear. Their 28nm asic wasnt due until late february, but is now delayed until May 1st. Customer support track record seems fair, but recent feedback from customers experience with their bitfury based gear is welcome.

Cointerra 3/5
Despite very high expectations, CT are over 1 month late, and missed both power and performance targets by almost 20%.  Supposedly just started shipping, Im not aware of any retail customers having received their gear yet, but should happen any moment.

Hash Fast 2/5
Still havent delivered anything yet in meaningful quantities. Currently >2 months overdue on their inititial shipping promise and causing a lot of controversy surrounding their MPP and refund policy. Facing litigation. Have shown working silicon, its assumed / hoped they will start shipping very soon, but confidence is this company is in free fall.

KnC 4/5
Have shipped several PH worth of 28nm miners and are the only company so far to have done that. Delivered (way) over spec and only marginally late on their first batch. High number of reports of defects and RMA's, especially on the first batch, but decent customer service. Currently no longer selling 28nm miners, only preorders for Neptune, a 20nm product with vague specs and vague delivery date (Q1/Q2). Optimists expect KnC to again vastly over deliver on their specs and achieve that on the early end of their shipping estimate, pessimists (like me) fear that while the specs are probably indeed understated, 20nm products are not likely to appear before summer. Based on their past achievement, which also exceeded my expectations, I wont deduct points for that though.

edit: One point that was raised by other members, and which is quite valid, is that KnC really dropped the ball on their hosting service. Prospective customers may want to think twice before trusting them with hosting.

Technobit 3/5
Doesnt produce asics, but produces and sells boards based on 3rd party ASICs like Avalon and Bitfury. Have shipped a lot of hardware already, but despite claims of ample stock, currently seem to take ~1 month for shipping. Also lots of complaints about their parcel service (speedy/DPD). Customer support seems to be reasonable though, even if stretched a bit thin.

Virtual Mining Corporation (VMC/AMC) ?/5
VMC is developing and selling preorders for 28nm structured asics, but havent shipped anything yet and havent really shown anything yet. Not much is known of their schedule and even less of the current project status, other than that its being delayed. Despite a public and verifiable partnership with eASIC, a well known semiconductor company with unquestionable expertise, and a business approach that in theory should enable fast time to market at relatively low risk, VMC is cloaked in so much secrecy that serious questions can be asked about their credibility.

edit: ActM announced a change in plans, their 28nm structured asic appears DOA, and instead a 55nm custom asic is promised. Probably too late to really matter, unless its one really efficient design.

I hope he does. Im taking bets its consuming well over 80W. Probably 100-120W.

I think a disclosure is needed in your post. Those that followed the AMT thread will know, but others may not, that you were sent this miner for free by AMT in an attempt to prove their legitimacy.  I think its also reasonable to mention that while you received your demo miner, actual paying customers still seem to be waiting for theirs for 6+ weeks now.

As for the machine itself, if you happen to have a kill-a-watt, Im sure many people would be interested in the results. AMT website claims 80W, Im curious if thats true.

Theoretically, yes.  But practically, we are so incredibly far removed from that, it doesnt really matter. Where exactly do you see a long term bottleneck?
If you think its assembly; globally we are currently producing ~50M less PC's than last year. Imagine even if only that (assumed idle) assembly capacity is brought to bitcoin miners, then the industry could assemble ~10 EH worth for Cointerra rigs per month based on idle capacity alone.

If you think its asic production, foundries currently have enough capacity for 300K 28nm wafer starts per month. TSMC recently said its running only at about 70% of capacity, so again if we only look at existing and idle capacity we are potentially talking about almost 100K wafers per month, interestingly that also works out to ~10 EH per month.

Its obvious we will hit economic limits long before we really run in to enduring production limits.

Only minor issue I see, its not clear to me how the PSU is supposed to breath. Is that BA logo plate really a grate, or is it breathing from the side of the case (not ideal for a rackmountable) or from inside the case, which is also not perfect.

X3 seems to have 6 fans. Either way, you are going to need active cooling.

X3 looks nice btw, one of the first bitcoin asic system designs that appears to make total sense (from cooling perspective).
And the fact they are underclocking their chips to achieve their promised power efficiency also implies that these machines should be able to overclock by 50%, provided you can supply the required power and keep them cool.

Yes it is, AMT kept making false claims about expecting to receive bitmine chips and being able to ship miners at impossible times longer after Bitmine made public their chips would be late and would not ship before January:
https://bitcointalk.org/index.php?topic=304605.msg4078302#msg4078302

Bitmine's statement was public, was repeatedly posted here, up to you to believe if AMT somehow "missed the news" and wasnt informed by Bitmine and didnt read posts in this thread or any Bitmine thread; or if they deliberately lied to their customers. Just like they continue to lie about power consumption claims and have been lying about delivery dates since the first day.

Astonishing how people will fork over their hard earned money just because there is a "sign on a building" and someone on the phone is civilized to prospective customers on some hours of the day.   Really? BFL have a sign on their building too, Im sure some of their people can be nice on the phone too at certain hours (at least if you havent ordered anything yet and arent asking about when they will ship or demanding a refund or RMA). Doesnt change the fact their core business is lying and swindling and selling preorders for vaporware.

Im yet to be convinced "alpha technologies" is for real, but even if they are, or someone else takes this on, the benefit over a GPU will be relatively marginal, both in price and power efficiency The problem with scrypt is that it requires massive IO speed and a decent amount of ram. GPU's (and FPGA') excel in IO speed, and a good memory controller is  one of the most difficulty things to design;  there is no hope a scrypt asic will do that better. They could be a bit more space and power efficient because they dont need all the functionality a modern GPU offers, but Im not very optimistic "amateurs" will be able to design an asic that is substantially better than AMD/nVidia GPU's for scrypt. For sure dont expect the quantum leap SHA asics offered, its not gonna happen. At most it will be comparable to GPU to FGPA transition, in so far we ever made that transition.