Topic: first bitmain antminer S10 speculation thread (probably) (Read 29509 times)

Numbers are good, but there are few ways to go. You can order Avalon miner, you can wait for 4 more months and wait to see is 16TH/S Dragonmint miner real(some weird things about there going on just use search), you can order from Baikal(buying through them is hell). Or you can order from Bitmain and wait for 2 months for your S9. Or you can just buy BTC and HODL.

Regarding the S10... As we can see in current moment how S9 works(and numbers associated with that) i don't think they will release it anytime soon(Maybe FAMOUS TULIP MINER that will crush Dragonmint! /s). I really liked ceratin topic that i have seen but can't find it right now. What was spoken there is "they secretly made S10 for themselves, and selling S9s as is) when S9s become obsolete they will start selling S10s.. But hey again this is just a thought(but i liked it )

I cant access bitmain too

So do you think it is still good time to buy s9 miner?

Thanks for good advice and explanations!


greetings

Nice reply, I'm with you Kamal on this.  You clearly know what you're talking about.  I have been in the ASIC business for 30 years and have done over a dozen tapeouts with TSMC. 

One of the few times a post is so spot-on I've quoted the entire thing
As you said, these smaller and smaller nodes and the processes of making them are targeting very complex LOW POWER applications. In most cases such as the Arm Cortex, entire Systems On a Chip (SOC's) that even running full-tilt consume less that 0.5 watts. That does not easily translate into a shitload of simple but densely-packed SHA cores on a chip sucking in over 10W...

Plus, as ya said: Diff would go through the roof. Be careful of what you wish for without thinking it through - you just might get it...

 You're probably going to die before a S10 model shows up - as has already been explained.

 S11 - likely to be a while unless Bitmain decides to do a "new model" on the same 14/16 nm node but on one of the "newer more efficient" process varients - for a fairly SMALL gain in efficiency.

I think a lot of people dont quite understand that while these large chip using companies buy lots of processing chips, they are wildly different thant a bitcoin ASIC. Fundamentally, all these companies looking to power portable hardware and are trying to make the least amount of heat and power draw while bitcoin asics are the polar opposite. Asic manufacturers are trying to stuff as much power into a small space as possible. Also, the asic market is tiny, a drop in the bucket compared to what these large chipmakers deal with. There is no real incentive for them to spend all that R&D on a niche market that wont give them much of a return.

Also, imagine if the big companies like samsung, intel, and apple did make a chip and mass produce it. The difficulty would spike like never before and all of you would be complaining that they are flooding the market just like you do with bitmain now.

I would think that by now Samsung and the other big firms would be making better chips but they are probably NOT going to put them on the market because it would seem like money laundering. I am over the s9. It's still profitable now but it will be a dinosaur by the time a new one comes out. I hope it will be 25-30 THS and hopefully, it will lower power consumption.

Shit, imagine what Apple could do (or possibly they have done) with their version of a bitcoin ASIC. It would probably be faster and with low power consumption.

Hope you packed for a long stay as nothing on the Horizon at all.

It's been awhile of no news about s10
anyone heard anything?

I waiting here to buy those s10 antminers

All the predictions and speculations were in the same S9 to 14 TH @ 100WxTH ...

Let's wait for the first semester of 2018, at least let's reach 10 nm.

I would really like bitmain to bring back a home miner

The world has run out of hosting and you cant run
S9's at home and stay married.

Why cant bitmain keep their promise from a few years
back and make new boards for the old S2's.

Today, I rummaged through the garage and dug
out an old S2, opened it up, blew it out and decided
to run 6 boards in it, to give more space between them.

PSU flashed up nicely, then decided to see how it ran
without the lid on and fans disconnected, it ran too hot.
So reconnected the two intake fans only and put the lid back
on.  It is running great, 50 degrees and very quite considering

Now 6 boards/blades with current chips would work
great, probably around 1TH per board, making this a
true home miner, and could do what the earlier miners
do, ie home heating, you could mine at home and
still remain married

I am sure there would be enough demand for the boards
and they should not be too hard to make, and they would
be cheaper to ship and you could buy a few at a time and
get around customs duties as well

Win, Win, Win all round

I see the news in a steemit, but is just a speculation, not an official announce.
https://steemit.com/especulacion/@btchispano/especulacion-antminer-s10
About the release date of antminer S10 I think need to await an official announcement. All news today are just speculations about official releasing date.

where did you see it?

I just read a spanish article about Ant S10 but without an official announcement. They say that in this time (Q3, Q4 2017) is posible to come out in marketplace antminer S10.

https://news.bitcoin.com/gmo-to-invest-10-billion-yen-in-its-own-7-nm-5-nm-and-3-5-nm-dedicated-bitcoin-mining-chips/

Looks like 7nm, 5nm and 3.5nm are all on the cards.

This could turn the mining industry from the hands of the many into the hands of the few. (not that were not heading in that direction anyway)

I also don't think there timeframe is realistic enough and dose not comply with moors law.

Buddy it is not been introduced to any place or any location. Please wait about it's releasing date and official announcement about the this information. I think this discussion is not necessary as of now. 
There was thread some days before in reddit to discuss about the same information but no detailed information there too.

I guess that we won't hear anything about next gen miners from Bitmain this year.

Very interesting this informations about the new Bitmain miner S10. Thanks for infos. I await the official informations.

You quote a year old post why?

There is no news because this is a useless speculation thread based on nothing but random guesses from people on the internet. There are no facts to support any of this.

Still waiting.... Is there anywhere to get some news? Not a lot of info from Bitmain these days

this is hard to imagine. I expect not that much increase in efficiency, due to semiconductor technologies you maybe reach 80mJ/TH, does some1 know? the chips are sourced with 3.3V ? I think so, they could find asic runnig at 1.8V, but i dont think this makes the consumption a half, maybe 3 quarter, then theyy could increase the quantity of units, and the density of units in one chip. this will become amazing, then will come the point you can not earn with bitoin because the currency cost is higher than the earnings.

Ok thanks for your information

When TSMC is talking about “high performance” they have e.g. GPUs in mind. NVIDIA and former ATI (now AMD) are TSMC customers. There are lots of high performance networking product e.g. from Broadcom or Qualcom too.
In principle it is everything, which is not battery powered and needs maximum compute performance.

Thanks for your info but about 7 nm nodes TSMC according to their website NOT ACCORDING TO ME has divided the 7nm segment into 2 divisions and on their website it specifically says one for mobile phones and the other for high computing power chips before that they have not divided this into 2 segments, may be these high computing power chips are for some other use also rather than only ASIC MINERS

No offence, but NotFuzzy is right.
If you want to build everything, which is currently mining again, from scratch, you would have to produce less than 60k 28nm wafers (assuming about 100 THash/s per 300mm wafer).
This is the monthly production capacity of a single giga fab. TSMC, GF, Samsung, Intel, UMC, … all the major foundries have at least one 28nm giga fab. There are not so much 14/16nm fabs around, but considering Moore’s law only 30k 14/16nm wafers would be needed in this case.
The theoretical revenue a foundry could make, if it produces all ASICs for 6 Exa-Hash/s in 28nm (or 14/16nm) would be less than $300M. That is about 1% of TSMC’s yearly revenue. I’m not saying, that this is nothing, it’s much more than I was predicting 3 years ago. But it’s probably still not enough to get any special conditions for early access to the currently very rare 7nm production capacities.

Thanks for the update about TSMC it was really informative

 Overall perhaps of ALL of their lines, but NOT on their 12/16nm node line(s).
 NVidia specifically moved to make their GTX 1050 and 1050ti at Samsung because TSMC didn't have the CAPACITY available to make those with all of it's then-current contracts (including the NVidia one for the 1060 and up).

 TSMC DOES have other lines still running on older nodes, like 28 and 40 nm (possibly 22nm but I'm not SURE there) that are kept busy enough to be profitable - AFAIK they still have at least one line running 55nm for older chips that don't NEED the current node but are still practical to use.
 In example, you don't waste the money making a radio/clock/alarm chip on current process because it doesn't NEED the ultimate in efficiency - same for Car "radio" stuff, boomboxes, televisions, and such.

Thanks for your reply I didn't said that bitmain is a bigger customer than APPLE AMD or NVIDIA of TSMC and for your information AMD fabricates it's chips from global foundry and not from TSMC Taiwan.
TSMC is running at 65 % capacity and they are taking mining segment seriously and for your information according to TSMC website they have stratified their chip fabrication into 2 divisions one is for mobile phones and the other for high computing power chips.

I think you know nothing about Bitmain mining machines and you are such a novice to mining machine segment and was not accepting this type of statement from a legendary member that bitmain only manufactures a couple of thousand miners each batch WHAT A JOKE
Bitmain has currently 3 active models S9,L3+ and D3  and if you look at bitmain Chinese website and have a common sense to use a Google translator you will know that in each batch their are around 5000 miners and with 3 to 4 batch released every month for each batch it makes around 45000 miners only for Chinese market and the same quantity  for international market it make around 80k to 90k miners and if you multiply it by 189 and 288 chips for some models you will get a hell lot a figure not as big as APPLE but a substantial number.

You obviously have absolutely no ties to the semiconductor mfg industry.

Ja Bitmain seems to make 'a lot' of miners but at only a couple thousand or so of them per batch even with 189 chips per-miner they are still strictly a boutique customer to TSMC. Apple, AMD and the other REAL major customers are each getting 10's to 100's of millions of chips per-month. Compared to that income from miner chips is pocket change.

As for when lower node miner chips will show up... Maybe mid-late next year if and that is a HUGE 'if' there is a need for it. These low nodes are geared to low-power devices where the effects of temps are manageable. You think running 16nm ASIC's at high power density is iffy -- forget about it at even lower nodes.

Bitmain purchases chips from TSMC foundry Taiwan and by the number of miner batch release dates now bitmain have become one of the major client of TSMC and they to  are taking mining segment very seriously
If you look at TSMC website for the 7 nm nodes segment they have divided that into 2 stratums one for mobile chips and the other for high computing power chips so they are taking mining segment seriously.
Now for the speed for the new model of S9++ or S11 (whatever they name it) we have to calculate the computing power for chips of newer nodes according to TSMC website they have started the tape out of nm 7 nodes in April and by the end of this year they will supply these 7 nm FinFET nodes to more than 10 customers and surely Bitmain is one of them.
Now come to speed of that 7nm miner we know that S9 speed is 14 ths with power consumption of 1372 watts
By looking at the 10 nm nodes efficiency according to TSMC website they are 15 % faster and consumes 35% less power so for a 1300 watt miner the speed would be 14×1.15×1.35 =21.73 Terahash for around 1372 watt miner

And if bitmain decides to skip 10 nm nodes and go straight for 7 nm nodes which I personally think they are planning and to raise funds for that bitmain is also issuing an IPO of 200 million $ in USA the speed of that ultra fast miner would be again according to TSMC website the 7nm nodes are 20% faster than 10 nm nodes and consumes 40% less power so the speed of that miner will be 21.73× 1.2 ×1.4 =36.50 Terahash and that to hypothetically

If it's a small increase it's probably not even worth it for Bitmain to create a new optimized chip at all with their low electrical rates and all. They're likely branching into LTC and X11 just to keep people interested in their brand.

There will never be a "S10". It will most likely be called S11 as their other models have been S1, S3, S5, S7, S9, etc. Although I don't expect a 10 nm chip anytime soon maybe they will increase the current chips efficiency slightly and jam a couple more hashing boards in there.

There are 800 coins to pick from.

And there are delisted coins.

I can find a list of every current coin.

But to consider picking 50 small value coins spending 100 usd on each. You need to find the list of delisted coins to fully know the odds of a three year move.

Ie. If there are 25 delisted coins and 800 current coins your idea may work.

But if there are hundreds of dead delisted coins your idea may be bad.


That is a very simple analysis

One big problem is you cannot foresee witch altcoin will skyrocket to the moon...
Besides, you have to pick a low to enter the market for it not to be too risky.
The middle approach is to mine AND to buy altcoin you think wil do great...

I was talking about a long time investment in altcoins with money intended for buying new mining equipment or a daily speculations on exchanges. As you know, during the last months some alts showed much greater return than BTC. I just wanted to know what experience had others in long term investing in a basket of altcoins - was it positive or not. I get that middle term investment (like a few months) in crypto currencies can get you hurt.

Power control cost.
Getting gear at decent prices

Lastly taking care of the gear so it lasts a long time.

If you can do the three above you can stand a chance to make money mining.

I have power deals with a few people that allow me to mine at profits.

And now most gear has stopped improving as it is smaller chipped

 Because if you set it up right, positive income is a LOT more certain from mining than from the very high risk of speculating.

 Don't take the last 5 months as "the norm", you have to look at more like the last 5 YEARS and the roller coaster pretty much all cryptocoins have been on.

Neither do I. What about not mining, but buying those altcoins instead of buying equipment for their mining?

 Global Foundries 14nm foundry is in New York State, it used to be an IBM foundry.
 They also have the ex-AMD foundry in Germany (Dresden I think?).

 Intel has at least one foundry in the US, Austin I think?

 There are NO foundries on the 14/16nm node in China that I am aware of, TSMC is in TAIWAN not in CHINA.

 Samsung has their 14nm foundry somewhere in Korea.



 The bulk of MINER manufacturing is done in China, but not the ASIC chips themselves.

Does anyone know or could point in the direction how these processes even work?... people like Nebulous Inc. who are still "in design phase" of the Obelisk claim to have a US ASIC manufacturer, its kinda hard to compete with CN semiconductors? Noo?

 The only folks I can see managing that would be BitFury (their chip WAS in production before Bitmain's was on the current node, but Bitfury sells to BIG farms and manufacturers only so it ended up under the radar for quite a while).

 Nobody else to date has managed to MATCH the S9 much less beat it - and I don't see any significant probability of that happening on the next node (very long shot if Intel, AMD, Samsung, or NVidia decided to enter the ASIC miner business, but I don't think that's likely as it's such a SMALL business compared to what any of those folks do NOW).



 The reason the L3+ isn't as close to saturation as the S9 is threefold.

 (1) S9 has been getting sold a LOT longer, there are a TON more of them in use.
 (2) The S9 has competition (Caanan/Avalon 721/741, whatever BW.COM has been using internally, Bitfury) that is fairly CLOSE to it's efficiency that has been selling for almost a year now (bit over perhaps) and HELPING it saturate the Bitcoin market.
 (3) The price rise of Litecoin during the last 4 months is over 10 to 1 - the comparable price rise of Bitcoin on a 10 to 1 ratio needed most of the last *2 years*, which has allowed a lot more time for folks to buy the hashrate up to the same ballpark as the price has gone up.

@ the yankees

yep..   I would look to make an asic better then the L3

as in terms of waste   the L3 is 4x better then the s-9

sha-256  is flatlined  for efficiency   at close to .11 watts a gh

with no major improvement.

This is good as BTC may just get more valuable as time goes on.


I see  much more alt coin growth to come.

the industry over all is over 124 billion

 I see  300 billion  in under 5 years.

So why build a better then an s-9  which is 20%  lost to power/maintenance

When the L3 is way better then the s-9 it is 5% lost to power/maintenance.

Not happening.

It simply is easier to do other coins then btc/sha 256

Then trade for btc.

So asic builders will target other coins.

BTC  is a big pile of money  in a wallet that can be 'mined' by putting in a password.

What do I mean?  16 mill  + coins are in wallets just under 5 million coins are in the 'ground'

So I mine ltc cheap and trade for btc
I mine x11 cheap and trade for btc

I have No incentive to mine btc at a high price.


look at LTC  only 5%  of coins are lost to maintenance

Sure, but competitors will not sleep forever (there are already quite a few good ASICs for altcoins) and may overtake the market of bitcoin miners, should they release a more powerful and economical miner for BTC sooner than Bitmain.

 They may not have AS HIGH of demand for the S9 as they did in the first 3 months, but they STILL sell out every batch in a day or two when they release one.
 Sure seems like the demand is still high.

Bring it in! I'll give you 10 bucks for it!

I seriously doubt we will EVER see an S10 model - Bitmain seems to have given up on rack-mount "huge power consumption" miner designs since their small-size single-tube type designs are ALREADY pushing so much power.

 S-11 will show up probably 6 months to a year after the "next semiconductor node" hits full production - be it 7nm or 10nm - as Bitmain isn't big enough to break into the early days and too much of the production of GF/TSMC/Samsung is earmarked to the BIG BOYS like Samsung, Apple, AMD, Nvidia, and IBM.

Why on earth would bitmain want to kill off their s-7 cloud mining on hashnest.

The answer is they don't and we will not see an s10 until hashnest s-7 goes to the red rather then firmly in the black.

At the monment the s-7 is at 45-50 percent power cost

Ie firmly in the black .

When the s-7. Is at 95 percent power cost they would look to upgrade to the s-10

No. This is one of those junk threads started by someone that doesnt understand how the industry works.

So, anymore speculation on this? Anyone hear anything new?

I need to know more......please advise if you dream a lot or there is some substance to your belief.

 Serious overkill. A 10MBit port will handle ANY miner for the forseeable future.
 1 Gigabit probably would be the installation, as those are common cheap and the norm any more.

 USB for a self-contained miner is dumb.
 Bitmain seems to have dropped the concept entirely since the U3, thank goodness!

ASIC on 16 and 10 nm will be sensitive to production. and too expensive. and do not forget to RMA case...bad with big machines..
 Bitmain avoid making large machine. for it is better to have smaller machines, and also because of the efficiency and recoveries of. and particularly money .
but the current situation calls already for bigger machines of higher performance in th/s becouse ROI..... price for BT is about 650, and that's not what we expected after halving.
If wants bitmain $ 2,000 per miner that must give at least 20 th/s. NOW

for the data port, i think either USB 3.1 or ethernet 10Gbit would be more reasonable, mainly because the convenience for the hobbyist, or for the noobs; for the ethernet, explicitly all of them consists of--- Do not forget to pay some of the license fees for USB and 10 gigabytes is unnecessary and expensive. It would be unnecessarily expensive already so expensive machines

If that is the final truth , also Qualcomm and Mediatek would not go for 10nm so early for their smartphone processors. They are doing this because they are expecting more performance and better efficiency (longer battery run time) at lower production costs (ideally). This drives the semiconductor industry since 50 years.

You will get for sure higher performance and better efficiency at 10nm compared to 16nm. But that the production costs are reduced in parallel is very questionable. This good old Moore's Law effect disappeared already at the 28nm to 16nm transition.
I bet, that Bitmain pays more for the pure 16nm silicon than they did for 28nm normalized to a TH/s.
And a Bitcoin ASIC is much more sensitive to production costs than a smartphone processor, especially if you have free or almost free power.

Well, Avalon's supposedly going to make a 16NM miner sooner or later, the Avalon 7. Not sure if that's true or not, but if Avalon actually make a miner capable of rivalling the S9, they might start researching the 10NM tech in hopes of making a more efficient miner. Bitfury's also purported to be making a 16nm miner, although those things cost your soul to buy. ASIC technology will get to 10nm sooner or later, but the 16NM process is still young and maturing; chips can only get more efficient even in the same gen.

10nm is not so far!

TSMC and Samsung already did several 10nm tape-outs (e.g. smartphone processors of Qualcomm and Mediatek). These kind of technology will be pretty sure part of the next gen smartphones in 2017.

Globalfoundries bought IBM's microelectronic manufacturing business completely in 2015. GF/IBM skips 10nm and goes directly to 7nm, but there will be for sure no 7nm products in 2017.

Last potential foundry player (in this league) is Intel. They are probably ahead of TSMC and Samsung, but very picky with respect of selecting customers.

Meanwhile Bitmain is a “midsize” TSMC customer. If they want, they probably can get a 10nm production slot. But why should they invest another $10M NRE if nobody is challenging them at 16nm?
Anyway I'm sure that Bitmain has already access to the TSMC 10nm technology and is designing the next gen ASIC. So they will go into production as soon as they need to.

I do not see an 10nm or and 8nm as IBM has developed in near future, not even in late 2017. Problem is the further down we go in nm, the less suppliers is possible.
So, too many big players will fill the orderbooks for other industries like the mining industry, so the S9`s will be here for quite some time. Perhaps they will release a 30TH miner that's able to run from 2 bitmain PSU´s or 1 IBM Bladecenter PSU, but a new miner like an S10 or an S11 is way out in the future. The competition on semiconductor industries has narrowed down with the 16nm and it will be even more difficult with 10 or 8nm.

Soon Bitmain will realize this problem and will either be forced to make the chips them self...not going to happen or they will simply stop selling to the public.
I guess the last scenario is the most possible one.

Bitmain's already skipped two rackmount miners - the S6 and S8; I really don't think we'll ever see a return of the monster racks that the S2 and S4 were ever again. They were never in very high demand, and the S3 and S5 were simply better options than the racks for the most part. -LN miners have probably been the replacement for the racks.

Bitmain does not have a high demand for S9, if they are thinking of bringing in an S10, they had better sell at a reasonable price. Miners are not making profit like before, buying an expensive rig is the last thing they want to do.

My professor and  a few of the students believe we have put togther a machine that can actually hash out 1.5 zeta hash

It is a bit tongue in cheek or dry humor. Just like my thread on the s-11.

Not sure about this topic. Trying to understand what people are saying.

My prediction is that if there is an S10, it will be a MEGA MONSTER miner - something on the order of A2 Terminator sized, or more likely 5-6U rackmount, with 60+TH and multiple power supplies needed to run it.

And it won't show up before late winter 2017.

if we're gonna speculate about a year from now, and call for 10nm, then I think 150J/TH is too much. Let's make some progress and expect 100J/TH or hey, let's expect 75J/TH.

75J/TH would make a 5TH/s home miner possible without melting the lamp cord extension cord that runs under the carpet to get to the miner.

To infinity and beyond!

Cheers,

- zed

so everyone is predicting with the antminer S9, right? even though S9 is ready to release, i am a bit curious about if the newer version of antminer is gonna release like S10 as mentioned. if nobody has had posted anything ever regarding the S10, then this will be the first ever thread.

if you wanna speculate with the S10, please leave some of the guess here.

i would guess about specification...

this miner will be released in the early Q2 2017, after the S9 is released.

for mining power, at least 15 Terahash/s at 150W each Terahash; because the production of chips will probably be done in Taiwan Semiconductor Manufacturing Corporation - TSMC (at my prediction, 10 nanometer).

why 15 terahash? because the transistors are ever increasing without enlarging physically, instead, by shrinking the fabrication of silicon. and for the 150W per TH, you know, a little amount of power can run a lot of things nowadays, and unlike 32 or even 16 nm.

for the data port, i think either USB 3.1 or ethernet 10Gbit would be more reasonable, mainly because the convenience for the hobbyist, or for the noobs; for the ethernet, explicitly all of them consists of.

indeed outlook will just like S4, i am not sure if S6/8 was ever released, if not released, then this thread is predicting if bitmain is bringing the S series ending with even number.

for the influence? this kind of miner will let the bitcoin network more fierce and more kind of altcoin will be gone popular as more people gets to mine, the bitcoin price will more likely soar up at least $1000. this kind of miner would gone insane in demand until a newer miner which could ever defeat this vorpal.

---

i am not sure if i am getting off-topic, but let's see what you guys think to this speculation.