Topic: DragonMint T1 16TH/S halongmining.com - page 40. (Read 87920 times)

trying to see how to do it. as I am willing to try to do it.

---

Moderator's note: This post was edited by frodocooper to remove a nested quote.

well on the first firmware the efficiency is poor

as tested on 3 different psu's  and 2 separate meters

there is a new firmware I will get to it in a bit.

.  Note I have  a  second demo coming  today according to dhl

so I will check firmware on the new one and do a direct comparison.

after 42 minutes  hash dropped to 15.68th

power is 1729 watts on the hp.

I do have multiple meters and it could be this particular meter is 50 watts on the hot side.
gui at 44 minutes

https://i.imgur.com/QXFwiBo.png


all tests  will be for 2 hours.  as power used does not shift a lot. 

So......so far every test I have seen have seen power efficiencies pretty much on par with or a little worst than the S9 14ths.

Think of it this way. At least the era of the Toaster S7s are coming to an end. At least the environmentalists will finally get off our backs!

---

Moderator's note: This post was edited by frodocooper to trim the quote from philipma1957.

You assume a lot. How do you know that none of the blocks solved by "unknown" pools aren't T1 pools we don't know about.  Do you take their word for it that it ONLY works on those two pools or does it ONLY work on those 2 PUBLICLY known pools?  Do we assume Halong has a farm somewhere?  Where are they mining on slush or a private one of their own??

You also assume that a client showing up with 1 or 2 or 5 million dollars to invest is treated the same as joe six pack home miner.  Believe me when I say that deals done at that high level generally speaking are negotiations and not just terms dictated on a website LOL.  When FOMO was in full swing the manufacture has the "hand" in negotiations, as the mining market shifts so does the "hand" during negotiations.

All in all your assumptions might be 100% accurate but equally as true is that a number of other "reasonable" arguments exist and therefor it seems impossible to accurately predict their sales volumes.

So it seems the efficiency is just barely on par with Bitmains 2 year old s9 tech. After all that bragging about it being the most efficient miner in the world, Halong is looking even more silly now.

to loud to sleep  it is back up

so far tests have been  not so good.

but I am now doing a third psu.

the myrig halong psu  did 1680
the delta 2400 psu   did 1610

the newest is an hp  2450  watt  https://www.ebay.com/itm/100V-240V-2450W-Power-Supply-for-Antminer-S9-T9-S7-L3-Mining-94-Percent-Efficie/122923683959?

just fired it up I also figured a way to keep it very cool.

gui has it at 15.98 th at 11 minute

https://i.imgur.com/dwgeUsZ.png

so far it seems to pull a lot of power   1610 to 1680 watts   and be just short of 16th on the better psus

the myrig  it did  15.5 after 2 hour test. and 1680 watts
the delta it did    15.7 after 2 hour test   1610 watts


the hp test is only 20 minutes  15.86th


you tube link of today's test

  note I will do this for 2 hours then I will do firmware then come back and test for 2 hours with new firm ware.

Your video will be live at: https://youtu.be/oR9fwKYIb-s

---

Moderator's note: This post was edited by frodocooper to remove a nested quote.

if you ever were at least one toe in mining operations whale biz you prob will know that usually one invites whale operations for demos way before the asic is announced in public, secure the contracts and later of course those guys get units first as well

who is this for example https://blockchair.com/bitcoin/address/3ACjzvyr3PdPCnAcAi2Nfw9RXcm9SVmtR9 mining with AB since mid March?

It is good reasoning actually. Think about it.
The batch 1 barely made a dent in slush. And slush is 3% of total hashrate or less?
Slush is 3.246 EH and btc difficulty is now 3.46T. I doubt batch 2 and 3 will do much either. We shall see I guess.

If they sold any decent amount, slush and CK should have huge increases because ONLY these 2 pools can take in dragonmints.
If thats the case, when they are delivered later, there should be a major increase in those 2 pools hashrate. The major pools are known as it is.

lol. They didnt sell to any enterprise companies I believe. I know the claims they make. Their telegram back then has like 600 people only then and many were toxic. No company will buy alot of untested units from some new and unknown company. No one hands over so much btc just like that. It is just stupid. Why would they be able to get into the industrial market like Bitfury when they cant even crack the smaller market lol?

Agree with u on investomer. Nice term

Please refer to this thread about SHA256 IC.  It has some discussion regarding ASICboost
https://bitcointalksearch.org/topic/m.33374953

If you want to google Scotten Jones at IC Knowledge and pull up his reports you will find all the information you need regarding process nodes and normalization.

Can you explain more about the difference?  The publicity is what got me thinking.  There was a huge bump in interest in bitcoin last winter in Korea, probably over the stability of their currency due to the DPRK...  You can buy bitcoins on in the subway terminals too!  In little plastic boxes!  

Are you speculating that the boost feature might only take up more space that could otherwise be used by more cores?  What is the point if it doesn't make anything faster?  As they proudly state on the front page of the web site, the designs are not based on simulations.  Maybe they never tried it without?  

I love the term “investomer”.  I think this is one of the peculiar aspects of Bitcoin mining equipment manufacturing.  Everyone sees a prototype and hype surrounding and they are quick to fund millions of dollars towards “production” well in advance essential giving quick payday to anyone who participated..  Yet, time and Time again that risk has played out very badly.  

Why not make true longterm investment into the development of the unit itself or startup and receive benefit for years off single investment?  Instead trust is placed in an unknown team with a prototype.  Why not put your trust in an outstanding team with wonderful reputation that could provide years of steady profits and access to equipment at cost? 95% of this thread and the risk could have been mitigated, but it sure makes for a great story and is a lot of fun

My experience was only related to the support and quality that they provide a-Si customers.  Koreans do everything at an extreme level and have a huge competitive spirit.  The USA invented auto-tune, Koreans took it to level 11.  I thought when Samsung got into SHA256 everyone else would be in trouble!  : )

I seriously was worried about my investment in Canaan 841s when I heard the news.  I didn't remember about BW.  It seemed like they were hyping this all of a sudden, so I assumed they felt something special.

Assuming it is true that the Innosilicon T1 uses Samsung chips I am truly confused with my continued support of Canaan.  

Samsung 10nm not equivalent to Intel 10nm.  Keep that in mind.  Samsung seems to have been using these labels loosely.  Also, Samsung has been aweful proud of making these ASICs.

There is also some question as to whether ASICboost routing actually increases performance.  It would seem that Bitmain would have implemented it if they thought it beneficial. Granted, it is early days with this tech and the software is being tuned.  Looking forward to seeing what the full potential is of these miners  is as everything gets ironed out.  So far so good.

That is specious reasoning at best...  They may very well have sold lots of gear and:
it's just not delivered yet.
it's still "burning in" on their own pool #BFL!
there sales were to mainly enterprise level companies who are running on "non-public" pools

With such limited information it seems impossible to accurately asses their sales volume.

I've been preaching about how bad a deal it is to be an "Investomer" since 2013.  It's not a popular opinion here in DEADWOOD.  The risk/reward ratio for being an investomer is soooooooo absolutely lopsided (and in case it actually needs to be mention it is NOT lopsided in the miners favour...) that I am constantly amazed how many people are so happy to line up to be one.

The only reason to not want to reveal identity is to LIMIT LIABILITY. No asic manufacturer or ANY other company out there hides identity. This has nothing to do with auto industry.
It is basic common sense.

I dont mean to be a wet blanket but there is no notable increase in hashrate in Slush and Ck pool. Slush and CK pools are the only pools that can take dragonmints.
Hashrate for slush should go through the roof it they sold alot, especially considering how small slush is to the total hashrate.
Furthermore, alot of slush and CK pool hash will be due to cheap S9s flooding market.
Thus, the logical conclusion is that they didnt sell much miners.

Halong's Terms say that all deposits are non-refundable. So, they can RUN AWAY if they want to and buyers have no recourse.
Halong's last post mention the high cost of making a miner and competing in this market as it is.

So, buyers should be careful when buying from this company. They may just dissapear.

If u really want this gear, I suggest diversifying and not putting all your money into this gear. Or split your buys. Buy more ONLY after a batch has been delivered physically to your home.
Chances are that they are not doing well. They may just dissapear suddenly and u have NO RECOURSE because they are invisible.
Their TERMS state that deposits are not refundable. Be safe.
It would be sad to see people being screwed should it occur.

---

Moderator's note: This post was edited by frodocooper to trim the quote from Dr.Mann.

Why?
The Foundry part of Samsung is no different than TSMC, or GloFo. They will make any chip for anyone with the money, in this case its SHA256 ASICs for Bitmain, Halong and Ebang. They used to make chips for BW, remember?

As for the 10nm eff... you need to pay attention to what folks here who are in the semiconductor business have been saying for almost a year - when it comes to mining chips the 10nm and lower nodes are pure and simple 'see what we did' Marketing with little bearing on possible real-world performance increases. The physics of these nodes strongly favors low-power (mobile applications) and as even 16/14nm has proven, pushing a high power application onto them gets very tricksy. Don't forget, Ebang's 10nm miner actually has slight worse spec than their previous miner...

Good numbers short term, not really. My S9 pulls 16-20Th for the first few mins. It just does not calculate properly at first. It will also take some time to stabilize the numbers.

http://ckpool.org/users/1JdC6Xg3ajT3rge3FgPNSYYFpmf53Vbtje
Pool side number is zero for 1 to 5 mins.
2.96 T for 1 day. Wat happened? U turned it off?

http://ckpool.org/#/user  . I added your address 1JdC6Xg3ajT3rge3FgPNSYYFpmf53Vbtje into that link.
Your last share was 8 hours ago.

---

Moderator's note: This post was edited by frodocooper to trim the quote from philipma1957.

please be sure to test the updated firmware as well. your miner should offer an update in the backend section.

---

Moderator's note: This post was edited by frodocooper to trim the quote from philipma1957.

So today is 11th of april, the daterange for T1 Batch 2 was 5-10 april - did ANYONE received a shipping number yet?

Why is it delayed once again and there's no information to anyone?

I was a little concerned when I heard Samsung was making SHA256 ASICS.  Hanguk mansae!

189 10nm chips per unit...  vs 104 16nm chips in the 841.

I expected more of an efficiency increase, especially from Korea...!  I hope it wasn't Samsung's fault.