Bitmain Antminer S6 and S7 Speculation

[sidehack]

Kinda like when BitFury talked big game about their 0.19W/GH chip in exactly one press release and never talked about it again?

[Jamphone]

My KNC confusion is pretty simple if they built this chip  and it 2 or 3x better then any chip ever built  why don't they really show proof of it. 

They could get a shit ton of investment money and then rob them too.

So far I have not seen clear evidence of a .1 watt mining machine mining as of today.

If they had the chip, they wouldn't be losing market share of the network hashrate distribution. KNC is in a lot of trouble. Those leaked financials from December 2014, at a higher BTC price, showed them bleeding value at a rate of 5% of their total book value ($4.6 million) every 6 weeks. That's an "bankruptcy on the horizon" rate. Hence why they needed $15 million of equity in January, but have nothing to show for it. They'll have to raise more cash again soon, even with a new chip. Production will cost tens of millions, and the factories in China will require at least partial payment up front.

While mining in Sweden sounded smart in 2013, the electricity rate is higher there than Iceland, Georgia, and the better rates of North America and China. Plus, their farm supposedly cost $20+ million to house less than 10 MW. I bet the Chinese spend far less than that to house 10MW. Throw in the cost of low pay scale employees, which in Sweden is $80K+ USD, 12 weeks vacation and benefits, the high tax rate in Sweden (25% VAT plus high income rates) and I would bet they're going to have to shutter that operation before the end of 2015. Throw in millions of liability in law suits. KNC may produce a chip, but the company is either going to have to sell it for nothing to a creditor/investor or sell the whole company. They simply do not have the funds on hand to manufacture a final product.

[philipma1957]

My KNC confusion is pretty simple if they built this chip  and it 2 or 3x better then any chip ever built  why don't they really show proof of it.  

They could get a shit ton of investment money and then rob them too.

So far I have not seen clear evidence of a .1 watt mining machine mining as of today.

If they had the chip, they wouldn't be losing market share of the network hashrate distribution. KNC is in a lot of trouble. Those leaked financials from December 2014, at a higher BTC price, showed them bleeding value at a rate of 5% of their total book value ($4.6 million) every 6 weeks. That's an "bankruptcy on the horizon" rate. Hence why they needed $15 million of equity in January, but have nothing to show for it. They'll have to raise more cash again soon, even with a new chip. Production will cost tens of millions, and the factories in China will require at least partial payment up front.

While mining in Sweden sounded smart in 2013, the electricity rate is higher there than Iceland, Georgia, and the better rates of North America and China. Plus, their farm supposedly cost $20+ million to house less than 10 MW. I bet the Chinese spend far less than that to house 10MW. Throw in the cost of low pay scale employees, which in Sweden is $80K+ USD, 12 weeks vacation and benefits, the high tax rate in Sweden (25% VAT plus high income rates) and I would bet they're going to have to shutter that operation before the end of 2015. Throw in millions of liability in law suits. KNC may produce a chip, but the company is either going to have to sell it for nothing to a creditor/investor or sell the whole company. They simply do not have the funds on hand to manufacture a final product.

 If they sold off the "0.1 watt chip"   Would be nice to get a lot of them.  

@sidehack how about 1000 or so  0.1 watt chips to build something with.  


 We can only dream.

[TheRealSteve]

Kinda like when BitFury talked big game about their 0.19W/GH chip in exactly one press release and never talked about it again?

Maybe a bit more like HashFast and Cointerra .. or even that ASIC|rising bunch.  But at least Spondoolies somewhat lives on and BitFury's expanding, so they have the lack of bankruptcy and skepticism of existence going for them

[sidehack]

Phil - I don't know enough about BitFury's current line to know if I want to bother working with it at all. I mean, SFARDS claims 0.19W/GH SHA performance and I know I don't want to play with theirs. If BitFury sticks to ~10W non-BGA chips instead of ~100W chips, sure that'd be awesome. From what I understand, KNC tends to like big fat stupid chips so unless someone else contracted the dev I probably wouldn't mess with 'em.

Spondoolies new chip, last I heard anything, was on the same size/power scale as the Rockerbox. I guess lines at the foundry are backing up because they're overdue for release but still in business. That's good. As much as I may like to rag on their product line as being overly complex and expensive, the economy does need a diverse set of manufacturers and they're one of the good (which is to say, not evil) ones.

[J0hnnyb]

any news from Bitmain/Hashnest?

they put 2000 "New" S5 for 420 US dollar to sell at the shop.

[alh]

any news from Bitmain/Hashnest?

they put 2000 "New" S5 for 420 US dollar to sell at the shop.

I assume by "New", they mean "Well tested for the last 3 months"? I guess the $420 price is in line with what I hear is a recent Bitmain price increase for the S5 (to $378).

Is this an odd way to "clear inventory" before a new product comes out? Usually a company will discount an older product prior to the introduction of a newer product (e.g. the S7).

[alh]

With all due respect, it's kinda hard not to laugh when "Lack of Bankruptcy" make a company worthy of serious consideration. 

Probably need to work on a new lexicon for Bitcoin mining:

New means "Not previously sold by the manufacturer and comes with a Warranty"

Doing Well means "Hasn't filed for Bankruptcy (Yet)"

[sloopy]

It is great to read all of your opinions on manufacturers especially when talking efficiency.

A point a couple of you made had me down a rabbit-hole of thoughts regarding the present and future of mining. So much that if you have the time I would enjoy appreciate understanding more about is the fact the S3 (S3+ or ++ most never saw, much less touched) was better built than the latest offering, the S5. Dual fans, metal case, routed airflow, ran much cooler and the S5 with plastic sides, no top, and no bottom, along with more speculation. It is probably unusual speculation, but I think has a real question.

More specifically to my area of curiosity is those of you who have been around and had hands on with every bitmain product available to you S1, S2, S3, S4, S5, USB models matter, what is your opinion regarding the next version of hardware?
point 1: which must already be running / being tested or 'burned in' by BITMAIN... today, correct or not and why?; has every model been burned in? How long do you think they are 'burned in'? I understand it takes some testing to feel good about a product shipping and I am curious.
B: How much market share does bitmain have versus bitfury including any speculation or known facts regarding other ventures, ie pools, private mines, trading, or any category you feel is applicable. Another side of this is simply of all the miners in the field today and 1 year from today, what are the numbers, your guesses please.
3: If, now seriously, if, there ever is a hashpower war / attack, or whatever you want to call it, (entertain the worst possible scenario from Mike Hearn regarding the money side of bitcoin facing off the hashing side of bitcoin) far out speculation like the bitcoin network is being attacked by the reptilian overlords from the future who use humans for cloning experiments and gambling...
...how much literal hashpower / which pools, people, companies, etc would have to collude hashpower to impact the network in the most-likely way? Could Bitfury do it, KNC, BITMAIN, F2pool and BITMAIN, how much hashpower does it take today assuming you have the knowledge? It has very much surprised me to hear how large the bitcoin network is in terms of hashing power and look at a mining pool chart and see so much hashing power controlled by so few people. How much hash could be held by the fewest number of poeople or could we ever see it all owned by one company? 
A for example on this one is say, Discus Fish / F2POOL and Kano.IS were being 'mind-controlled' by those future lizards and the lizards can only impact the people owning / running these two pools, what could they do with current gen, and do they have anything on the workbench of R&D lab which could own the bitcoin network in a negative, or positive way, and what is that way?


 

[camsarria]

Www.bitcoinbrothers.de

Winter is comming! August 1

[soy]

So, a July miner from Bitmain hasn't happened so far.

[SunnyIgor]

I think it was a ""end of July miner"" from Bitmain
 

[sidehack]

Yeah, there is still quite a bit of July left. Bitmain likes to announce a couple weeks before they actually start shipping, but a July announcement for an early August miner would still be okay.

[philipma1957]

Yeah, there is still quite a bit of July left. Bitmain likes to announce a couple weeks before they actually start shipping, but a July announcement for an early August miner would still be okay.

Be interesting to see what the s-7 does vs the s-5. 

will .52 watts per gh drop to .35 watts per gh.

My guess is .38-.41 watts per gh for the s-7

[sidehack]

A 2x16 string on the S5 instead of a 2x15 would have made it a 1TH miner at 0.43W/GH wall. I'm assuming they'll have a new chip that does better than that. I'm rooting for board-level top clock at or below 0.32W/GH, but that's not based on anything just what I'd like to see.

[notlist3d]

A 2x16 string on the S5 instead of a 2x15 would have made it a 1TH miner at 0.43W/GH wall. I'm assuming they'll have a new chip that does better than that. I'm rooting for board-level top clock at or below 0.32W/GH, but that's not based on anything just what I'd like to see.

I'm hoping for a new chip.   Your thread has surprised me at what the current chip can do.   They really could make a next gen and make it more efficient.

Anything in low .3W/GH would be nice.   Maybe a underclock option to get in high .2W/GH would be even nicer.

[sidehack]

If they keep going with the unregulated string option (which does offer the best efficiency and cost) it'll always run at a fixed operating point, which for chip cost reasons they'll always set near the top of the chip's practical limits. They could make a 0.38W/GH wall 1.4TH miner in the S[odd] chassis with existing chips if they wanted to, but the chip cost would be about twice that of the S5. I was surprised they didn't do something like that for a rackable miner, instead going with BM1382 for the S4+ (where the BM1384 could have given 4TH off the same power consumption and the same basic design). Maybe they'll put out an S6 with BM1384 in a high-density string and an S7 with the new chip, and charge a premium for both because they're both pretty good gear.

[QuintLeo]

Bitmain's site is now claiming to be out of Batch 5 on the S5. I suspect they now have incentive to get around to releasing whatever is next, since they appear to have managed to clean out their old stuff despite their price gouging price raises this month.

 (edit) Never mind, it seems to be temporarily "out of stock" when they're adjusting the amount of price gouge. It's been bouncing in and out for a few days now.

[philipma1957]

Bitmain's site is now claiming to be out of Batch 5 on the S5. I suspect they now have incentive to get around to releasing whatever is next, since they appear to have managed to clean out their old stuff despite their price gouging price raises this month.

No that was merely temporary while they raised prices on it from 399 to 414 plus shipping.

They have no reason to sell s-7's at this moment.  The current pricing is working and they will continue to do this as long as they can.

[jonnybravo0311]

Wasn't the design of the S2 supposed to be that modular machine?  A bunch of boards on a backplane that could easily be swapped out.  Unfortunately, they (Bitmain) never followed through with any upgrade kits.  That kind of setup would be great for the datacenter guys.  Keep the chassis and just upgrade the boards.