When I saw the steep increase on Avalon's batch #3 price compared to batch #2 I was upset, and I had a public argument with Yifu. Here go a couple of excerpts:
Love when people throw around numbers like "3 month" randomly without any math.
the current difficulty is a little below 5,000,000, which a three module Avalon will produce ~6 BTC a day under these conditions.
Let's see how "3 month" ROI actually looks like.
If difficulty raises to 30,000,000 each three module Avalon will produce 1 BTC a day, enabling a break-even in 90 days.
So yes, if you think we are going to get a 6x difficulty increase between now and May, let's even say June! then by all means, please don't purchase Avalon, because I wouldn't.
I just don't think that difficulty will be 10,000,000 when the batch #3 units will reach the customers. Two questions:
- can you guarantee batch #3 delivery not later than May?
- wouldn't the difficulty reach at least 15,000,000 once all your 1,500 units are deployed? Wouldn't be much higher if ASICminer/BFL customers also start deploying more units?
- Yes, but I think this question can be answered based on how fast we ship batch #2.
- not really, the current diff already includes majority of batch #1's hashing power. but I also don't believe diff will
only be 10,000,000 but like I said, it should take at least 1 month of ROI, realistically best case 1 month, regular projection of 3 month, which is about diff 30,000,000 which is more reasonable
if BFL ships.
Oh also, expect a newsletter soon.
I thought: OK, we will get our unit in May, thus a 10/12 million difficulty is reasonable. You will remember that Yifu set the batch #3 price doubling the difficulty we had at that time (slightly less than 5 million), which gave a 30 days ROI for a batch #3 unit priced at 75
BTC (Avalon expected difficulty to be around 10 million when we were going to receive our batch #3 units). I accepted that explanation as an honest proposal. Avalon was disappointed by the fact that some of his batch #1 customers were selling their units on Ebay with a huge mark-up, while Bitsyncom was probably still in red, so they made batch #3 customers to pay the profits for the other batches too. That was very OK to me if the units were delivered in May as promised. One month delay is acceptable too in this crazy ASIC world. But I'm afraid batch #3 customers still have a looooong way before receiving their units (July? August perhaps?), thus I'm afraid that ROI is gone for good.
Unfortunately everything happened upside down: batch #1 customers looked at the biggest potential profit, but they were charged less than any other customer because nobody was sure if Avalon was going to be able to deliver ASIC. As soon as they proved to be legit, batch #3 customers were charged a very hefty price that made ROI possible only if all the promises (delivery date basically) were kept, but they were not - therefore, batch #3 customers will very likely mine at a loss.
And by the way, difficulty is going to go x6 between end of March (when the above argument was held) and June (see bolded part). I would say it's not Yifu's fault, in fact I admire his work for having delivered the first ASICs to consumer hands, and that was greed (not his greed, but everybody's greed in this market) what set Avalon's batch #3 price tag. Nevertheless, we are approaching very fast the point in which mining will be a very competitive endeavour, where ROI is only achieved by saving every penny on electricity/storage, and where a 5% decrease in the exchange rate really shakes the miner's business plan. Making a long story short, we're going back very quickly to the same point we had just a few months ago (pre-bubble) with GPUs.
