Topic: Bitmain BM1385 Chip ( 230W / T) & AntMiner S7 Discussion Thread (Read 5091 times)

If they go with a top-clock strategy like they've done on the odd miners so far, and keep an unregulated string like they seem to prefer, we'd be looking at a 2x17 string allowing 2.55TH per machine but at almost 700W. If for some reason they did a single-wide chain instead of two chips wide (which is a lot more likely to have balance and stability issues) we could see a 1.3TH 350W machine but that's not terribly likely.

Or they could change things up and go with VRMs, lose 10% of their power and still actually be about the best thing going. Problem is, if they want to use the TPS53355 like almost literally everything uses, they're looking at a VRM per every two chips, which pretty much doubles the cost. Using a multiphase setup like I believe Avalon pulled out for the 4.1 might do a better job but still greatly increases the cost. It's possible they'll build a miner under 500W but there's a lot of hindrances to doing so.

The crash happened BEFORE Bitmain announced the BM1385.

 Most of us that have published estimates are estimating that the S7 will have a little less power draw than the S7, at a hair more than double the hash rate.
 I don't think anyone has published a serious estimate under 500 watts, the concensus seems to be 520-550 watt ballpark.

First to go completely off-topic I respectfully request Dogie to please stop moderating or at least publicly. If you wish to report threads of course it is your choice. Regardless of fault, every time you do this it starts a flame war. I hope you can see it doesn't matter how right you think you are in doing such it causes the rest of us to wade through the crap and then usually the back and forth. Again, it doesn't matter who is right or wrong, it is a reality we must go through to get to what we are here to see.

I will go further and state in my opinion the information you offer in threads outside your review threads lacks research and experience. In your review threads you are normally consolidating information from the manufacturer and have good resolution pictures with the basic information one might need to reference from time to time. It is what many n00bs see for the first time when searching out a bitcoin miner. It is perfect for n00bs to get a feel for the ride they may take, to find the miner on the network, remember the power specification, etc. but for most experienced miners with basic electrical knowledge it is a repository of information. Do not get me wrong, it is good to have such and I have searched your name just to grab a link from your sig and look up a specification.

At some point you must realize the public attempts at moderation and reply to every topic with what has unfortunately many times been inaccurate completely overshadows the good things you do contribute with your review threads and again, this incites some of the very behaviour you report to the mods. It does not make this forum a better place. I am not convinced it is a maturity issue, a need for public confirmation, or you want to be a moderator so bad you are trying to show you can do such. The why doesn't matter except your actions push me to believe you are actively pursuing one or more of those things.

Rather than taking my comments as an attack and reporting it, give it consideration and you do not even need to reply. I think you have heard this all before from others.

  The moderators do a fine job here of straddling the lines. I hope your private actions are helpful to them at times. I do not think it is too much to ask that you let them do their jobs and keep your forum moderating recommendations to private communications.

I am confident some replies to my post may include it is my choice to come here, I can use the ignore button, it is everyone else's fault, you have a right to your opinion, and several more items which are not applicable or plainly do not matter when the core issue is, no one wants to read your opinion regarding the thread being locked especially when it does not need to happen.

This thread is / was fine and while I do not wish to debate the validity of this thread I do not have any intentions in continuing a conversation regarding this request to you so I will weigh in with my own opinion. In no way should this thread have caused you any inconvenience. Your comments regarding deleting it may influence a busy moderator to hastily delete it. Good information, opinions, and experience are tremendously hard to come by. I want to read the opinions of several people who have posted in this thread. Through prior actions and shared thoughts some of these people have shown me I want to read their views on any topic and they have experience and knowledge which is a huge asset to this community.

So please try not to take this personally and try to view it as constructive. I know of several people who appreciate your guides and I am one of them. Lets try to build from that. When you do the public moderating and replies to threads with bad information it harms the very thing you are after.

Let the flames begin towards me from any and all, but I sincerely have zero intentions to read or reply. I am not out to be a public figure of bitcointalk, simply a regular reader and follower of honest, experienced, knowledgeable folks who know their limits and maybe a couple who are funny enough the rest doesn't matter  

On Topic:
Obviously I am not pekatete but I did want to reply to anyone thinking the bitmain announcement had an impact on price.
From what I have read on the web and in general social media it is / was the flash crash fire sale which caused the initial push. Whether that has been the sole reason the price did not completely rebound, or the price was even slowly being pushed back down... hmm who can say for sure? I certainly think the bitcoin core vs bitcoin XT vs notbitcoinxt is hurting us. If anyone has 10 grand in their hand and opens a website to evaluate bitcoin they are going to see an article about this debate. I do not think it plays out well in the media because it isn't a good thing.

Do not take me saying it isn't a good thing to mean I am against XT, the block size increase, etc because that is not true. I am against the way this is being handled by all of them.
What I am most against was pulling the curtain back and showing the world dirty laundry.
It is akin to Dorothy seeing Oz behind the curtain pulling levers and smashing buttons to make the fire and smoke appear to be a God / Wizard, etc and once you see such, you do not unsee it.

Bitcoin has now been presented to the world as a store of value controlled by a group of programmers who cannot get their shit together and by the way also control the course of bitcoin. True or not, that is what has been presented for the uninitiated bitcoin user to consume.

Is that what they wanted? Seriously? I blame all of them, and that isn't going to make me any friends but I do not believe this was the only choice. I do believe this is affecting the price and will continue to do so. I think they have damaged our landscape and only time will tell how bad the damage will be, and if it is temporary or permanent.

I have read, watched and listened to everything I can find on the subjects of block size, XT, Gavin and Hearn's history along with most of the others with current commit access and I learned some things. Mainly that Gavin seems to be a bit easy to manipulate (by the right people) but has a good heart and the right intentions.

The core devs should have been way more vocal to the media and met Gavin and Mike head-on in the bitcoin arena with sound counter proposals. This should have occurred well before Mike and Gavin had any momentum. I believe it was the inaction and refusal to simply allow a temporary increase and if they'd all signed on for a 1mb to 2mb, (or MB whichever in this context) then they would have breathing room so they could all argue about the specifics of increasing fees for transactions, etc, and on.

Personally, I am a miner, regardless of my farm being tiny, small, or large by home miner standards, I pay for everything and don't get anything for free and I think that does give me a right to my opinion just like it gives anyone holding some coin a right to voice theirs. My opinion is the core devs, Mike, and Gavin are all at fault and of all the choices which could be made we are on track for what is bad for bitcoin from a perception standpoint.

We must remember that reality is perception is reality, and the way many will perceive this debate is bitcoin is already centralized due to being affected by a couple of guys who are part of the programming team who make the bitcoin code. If those two guys have the power to cause havoc then it certainly isn't the "decentralized", "no one controls", "government cannot harm" "money of the internet" it has been billed as by the bitcoin community.

I have done my homework and understand most of what has transpired. I understand Mike and Gavin do not control bitcoin, nor do the core devs. The reason the price is being impacted is because  the perception of many people who may consider investing or who are hoarding and will (have?) start dropping coin.

I do not think the recent flash crash was caused by XT, but the depression of price is impacted by XT and it very well may get much worse before it gets better. However, I am not turning any miners off, I actually added quite a few TH and picked up coin on the low side. I recommend we all invest a little more of what we can afford to lose and keep doing what we do.

Well, it's not like they lay out the masks by hand. Either they automate the design of masks from RTL or HDL, or they invoke a cell library that the foundry supports, that contains elements that are used in the construction of this design (most likely just logic, some storage, some lookup tables, and some clocking elements)

if they didn't use a library... Then this seems they would have very top engineering to produce possibly the most efficient path?

No, that's not what it means. It means that the chip's production and design was fully customized. Although I'm not aware of the exact details, I'd assume the actual design from the ground up was customized (either invoking an ASIC cell library or possibly not), and possibly the production process (although I doubt that more).

Whatever effect it might have had would have been lost in the general price drop due to the 'fork' hubbub, and somebody dumping Bitcoin on Bitfinex; https://www.reddit.com/r/Bitcoin/comments/3hiwam/everybody_settle_down_the_price_crash_has_nothing/
don tinfoil hat

plz peka, explain better, why this news,
has had as a result the fall in the price?

really interested!!!

Part of that would require going back below 400W power dissipation on those heatsinks, which guesses for the S7 put it at 500W again like the S5.

Nice, I can't wait to get my hands on the S7! Congrats on the impressive power efficiency, even at 28nm. 

Bitmain, is there any way you can adopt a lower airflow design on the S7, like the one used in the S1 and S3?  The jet engine included with the S5 is a little too intense IMHO.

Looking forward to the specs!!

https://bitcoinmagazine.com/21660/bitmain-announces-launch-bm1385-chip-future-antminer-s7-bitcoin-miner-increased-power-efficiency/

In need of new rigs, come on bitmain spondoolies, get some new fancy products for christmas! 

http://www.itop-corp.com/asset/cache/__h64__/internal/logo.png

Thank you very much for clear explaination .

It means that the entire logical and physical layouts are done from scratch without using pre-made IP blocks. Most semiconductor fabs have built up libraries of various logical functions to use in copy-paste design.

The problem with using pre-made or 3rd-party IP blocks is that they are optimized to perform their specific function but not necessarily optimal when it come to working with other sections. Even worse is that they are stand alone islands that must be connected to each other and that leads to signal routing problems and performance hits from excessively long connections between the blocks.

With full custom everything is done from scratch using pre-made IP as perhaps a guide to what needs to be done in each block but that is it. Given Bitmain's experience with their previous chips they know what the logical structures need to be, now it was just a matter of translating those structures into optimized pathways on silicon. All physical layout of the chips is done with point-1 being shortest possible connections inside of and between the various bits like memory, math processors, coms, etc. Shorten the internal connections and switching losses drop like a stone allowing faster speeds and/or lower power.

ASICMiner's BE300 proved eight months ago that there was still substantial room for improvement in 28nm.

Get over it.
By and large, the '14nm' tag has become a marketing tool much like IoT. There are few devices that can benefit from going there (mostly phones, tablets, etc.) and there are devices that just-don't-need-it. As I and others here have repeatedly said, node size is NOT a major factor for mining ASICs. Bitmains new chip just proved that existing mature and well understood node sizes such as 28 and even 20nm have lots of room for performance improvement vs using 16-14nm with its inherent risks, mask costs and poor chip yields.

Is LK announcing the smaller process die chips, or are they still sticking with the 28nm finfet?  I swear I read somewhere LK was pushing 14nm.  Ah yes... here's that thread: https://bitcointalksearch.org/topic/lk-group-limited-will-start-bitcoin-miner-14nm-asic-project-1033676

Well, this is good news. But with this efficiency I doubt it will be suitable for home mining as the noise could be too much. It's good that we can mod s5 fans, I hope we'll have the same chance. I know the last thing Bitmain worries is the noise levels, but it could be cool if they just work on it a bit.

Maybe they're also hoping to steal some thunder from LK Group, who I understand are to announce their new gear intentions on the 20th - which for China would probably mean in about 12 hours.

There was a flash crash apparently. Someone sold 2.3M at Bitfinex.