Topic: DragonMint T1 16TH/S halongmining.com - page 37. (Read 87776 times)

It is a good graphic easy to see if things are working close to correct.

Lol. This gave me a chuckle.

I'd still have a drink with the Dr even though we don't see eye to eye. I look forward to him uncovering Halong's offshore legal service address in Belize

Looks good! I like the visual specific chain tracking. I'd like some specific tuning capability soon!

I downloaded the  file and did one unit. firmware upgrade
https://i.imgur.com/dbPuN0M.png
https://i.imgur.com/5uSEDSq.png
https://i.imgur.com/bL9ES0t.png
https://i.imgur.com/kmQFdl6.png
https://i.imgur.com/6SBTNuu.png

@Sandal_Hat: I actually can't believe I have to defend myself here for doing the exact opposite of what you claim. I must insist that you do a basic fact check before you accuse anyone of "unethical conduct." You could not be any more mistaken about me and Unicornflex. Contrary to your claim, I have never praised Halong Mining for its greatness or otherwise tried to elevate the company for "marketing purposes." Unicornflex never did so either. I am not shilling for Halong or anyone else. I have no secret PR agenda or ulterior motives. Unicornflex has done the exact opposite of what you have claimed. The link you posted in which you allege he is "praising" Halong, he actually does the opposite and excoriates Halong. He made the case on why he has doubts about Halong, and in a constructive manner urged Halong to set the record straight, which Halong declined to do. I have written in detail providing reasoned support for the validity of his concerns. I did so because I shared his concerns. In doing so, I have painstakingly tried to be as objective as possible and provide as much supporting evidence as possible.

The "massive" sales numbers of the DragonMint that he and I shared with this community were intended to show not how "great" Halong Mining is, but rather to demonstrate the potential magnitude of the epic scam that he and I both suspected that Halong Mining could be. I did not share this information for personal gain, but rather to try to help people possibly avoid getting scammed. I have no other motivation. At the time I first started posting in February, I was all but convinced Halong was a scam. I have since written and still believe that we should all consider Halong Mining an illegitimate operation until and unless it reveals its incorporation number and proof that its claimed FCC certification for the DragonMint actually exists. After reputable individuals vouched for the existence of the DragonMint, such as -ck, then and only then, I posted that I felt I would receive the DragonMints for which I had paid, but I still maintained that Halong was not behaving in the way legitimate companies operate, and that all potential buyers should beware.

I know Unicornflex only because I bought some Antminer S9 units from him last year, a transaction that I also shared using the "Trust" feature of this community. I have no interest in promoting or questioning any hardware mining company. I am an actual customer of Halong Mining who demands transparency from them. Support my efforts in that endeavor for our mutual benefit, or don't. Your choice. In my case, I regretted buying from Halong once I realized I could not verify Halong's incorporation. I still regret my purchase. I became even more worried once I learned Halong was unwilling to do anything to help quell fears about claims Halong was a scam operation.

To help others make informed an informed decision about Halong, I shared the results of my due diligence research on Halong and its affiliate "MyRig, Inc." Because others seemed to value my research (into which I put considerable effort), I continued to post in this thread to help shed light about Halong and MyRig and to help introduce myself to this community, which I have come to appreciate. We have seen progress. In response to my posts and those of others, this community's collective examination helped to reveal the Japanese incorporation number of "MyRig" (at least its name, apparently translated into Japanese) for the first time in this forum. In addition, I revealed that MyRig, Inc. is not incorporated in Colorado even though it holds itself out as authorized to do business in that state. My research also supported the inescapable fact that Halong is an anonymous operation that is attempting to operate with impunity.

I shared in this thread numerous times my skepticism of Halong Mining and my concern that I would not receive the DragonMint units I purchased. I explained I felt this way because Halong miserably failed my due diligence tests. I posted that I thought my buying from Halong was impulsive and a mistake on my part. I vowed I'd never purchase from Halong again. (I can prove I am a Halong customer through photos and video if I ever receive my April Batch 2 shipment.)

I do not care about the success or failure of any mining hardware manufacturer. I have no financial interest in any of these companies, including Halong and Bitmain, except to the extent I may need warranty work from one of them some day on my miners. However, if a company operates ethically in the marketplace, delivers good value, gives good support, and treats its customers with respect, I hope it succeeds. Halong has not operated this way for the many reasons I have discussed at length previously.

I've tried to be respectful (yet fair and firm) to everyone and every business in this thread, including you, Halong Mining, Bitmain, and others. I've been as objective as I can possibly be. I've fact checked extensively the subject matter about which I've posted.  I've tried to support every statement I make with sound reasoning or hard evidence (for example, screenshots of the FCC database when I could not verify Halong's FCC certification claim, or Secretary of State records when I could not verify its affiliate MyRig, Inc.'s incorporation).  When someone posted that they doubted one could sue Bitmain, I wrote extensively explaining in detail exactly how you could sue Bitmain anywhere in the world.

You are off base suggesting that I have questionable ethics. I've behaved above board. I did not suddenly discover Unicornflex's post, I was speaking with him the day he posted it back in February. I disclosed this fact in an earlier post. I have not posted yet on any other threads in this forum, because I am a new here and want to wait to make sure I post only when I have value to add. I understood that was the expected behavior of users of this forum. Do you realize that? I would encourage you to do the same instead instead of acting in an unacceptable way toward new users. Judge newbies by the content of their message, and not by the creation date of their account and the number of threads in which they post. This community should welcome newbies to make constructive, objective contributions -- and not verbally assault them as you have done.

I regret if some find this message too personal or "off topic", but being publicly accused of acting unethically when I have taken the utmost care to act honorably, as Sandal_Hat has done, is way beyond the pale and required my public response. Thank you.

what is the optimal range number ?


Miner Info

#   Hash Rate   Status---Accepted/Rejected------   HW---   Temperature
1   5.25 TH/s   Alive-------1027/0------------   617---   71 ℃
2   5.27 TH/s   Alive-------1107/0------------   864---   68 ℃
3   5.15 TH/s   Alive-------1028/0------------   1215--   69 ℃

these are the 3 temps for each board.

I posted a review of the unit I received here, with tons of hardware and GUI pictures

https://bitcointalksearch.org/topic/reviewing-the-halong-mining-dragonmint-t1-a-10nm-bitcoin-miner-3311505

Tuning kicks in when the temperature is in the optimal range.

I'm kinda afraid that -ck may be politically or contractually restrained from answering my question. The only answer that he could make would be "feature parity with competition".

I used to work in the medical business and that phrase would actually mean "misfeature parity with influential political lobby". Imagine an electronic sphygmomanometer which instead giving you a numeric answer like "120/80" would be restricted to giving only a binary answer: "You are OK" and "Go see a doctor". ( This isn't a real example, I just wanted to write something that nearly anyone can relate to, without going into explaining medical procedures. ) In the medical field doctors frequently won't give you straight answers, to get to the truth you'll have to interrogate their office staff, both the medical and administrative side.

if you check my 4 dragon mint t1 miners at the solar array 1 board dropped

http://ckpool.org/users/16yLHLoeyuCLPMXkVpC3gyrRYvwRGwjKJr

worker .3

Most likely clock was too high.

When the 5th one comes today I will bring it to the array and reboot that  other one.

so 4 should do 64th

my 4 are doing


"hashrate1m": "56.2T",
 "hashrate5m": "56.3T",
 "hashrate1hr": "55.9T",

1 board is off  so that is 5.33th  so 55.9 + 5.33 = 61.23  of 64 = 95.67%  of spec hash  assuming a reboot fixes the no hash board.

they should do 4 x 1480 = 5920 watts

so 64th = 61.23 th

5920 watts = 6480

btw this over watts is a pain for me as I have 30 amp pdu's  and 4 of these are doing 28 amps which means  3 per pdu not 4

I have a fix  3 T1's and 1 L3+  pull 23 to 24 amps.

Precisely correct. The user can set a global Vcore and from there the tuning mechanism changes chip freq to keep what in analog terms would be called the best S/N ratio. Somewhere in Canaan's docs they describe what is checked, along with temps as I recall it is a mix of tracking HW errors and/or invalids then adjusting freq to keep them within their boundary values.

No matter what, it beats the hell out of doing a 1-time tuning run after bootup as Halong does.

It is probably only partly true. They may tune the clock frequency synthesizer per each chip, but I think that the hashing core voltage is getting tuned in groups/strings/whatever is the actual limitation of their voltage regulators.

Since I was given  the demos  from MyRig  total of 5x T1's  and 1x B52  along with 6 psu's

And buysolar and I will be involved with Canaan repair and upgrades on east coast of USA.

We will be in a very good position to directly compare the gear and the service. Of both companies.

Frankly I want to do business with both companies and continue to do so.
I also want to be honest  and transparent to bitcointalk.org community.

I do not have enough info to say buy canaan don't buy bitmain or Halong dragonmint.

I do have info info to say :

First)      Dragonmint T1 is real

Second)  Yoshi did do as he promised to do for me in fact  he did better as he also sent  psu's and  the dragonmint b52

Third)     The  Dragonmint T1 is a bit short of spec

Fourth)   The Dragonmint b52 is a higher hashrate then spec and higher power use.

Fifth)   If the Dragonmint T1  did 17th at 1610 watts  people would be happy

Note 4 and 5 are kind of linked in my mind

@ frodocooper  please leave in  number 4  + 5  as  it shows that halong does make at least one miner at spec

---

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

Ja.
At least Canaan seems to have done tuning right. Per their info, each chip constantly is being tuned and tweaked for best performance so as conditions change so do the chips speed.

Heat sink design for maximum dissipation/smallest package and lowest noise is another puzzlement regarding the Halong miners and Bitmains. Once again, Canaan put some thought into it and did it right.

---

Moderator's note: This post was edited by frodocooper to correct erroneous quote attribution.

All the longer term people myself included  are getting  15.10-15.50 th  and  high power draw at kwatt meters 1610-1679 watts


here are 4 miners  all dragonmint T1

http://ckpool.org/#/miners

10   16yLHLoeyuCLPMXkVpC3gyrRYvwRGwjKJr   52.6TH/s   6.86TH/s   1.06TH/s   4   151,513,653   71926881   99%   149,647,249   0.00010664

5th one comes today.  yet to do firmware.  and with buysolar  going to China  for avalon training.  WE  won't be doing firmware upgrades for 10-12 days.

http://ckpool.org/users/16yLHLoeyuCLPMXkVpC3gyrRYvwRGwjKJr

---

Moderator's note: This post was edited by frodocooper to remove unnecessary quotes.

Hey man, nice angry rant. I clearly stroked a well known problem area.

Coin mining isn't a very complex problem. I'm pretty sure that an average car enthusiast is familiar with how modern cars autotune their engine management unit and automatic transmission shifting points. And how those are in need of recalibration when the car's driver changes for one with a different foot temperament or the driver decides to switch to a different brand of fuel. Even after a service or some part replacements cars need time to retune themselves.

But cars aren't typically operated 24*7, whereas coin miners are designed for continuous operation. The actual probing in the tuning process doesn't need to run for multiple tens of minutes. It should take just a few seconds, maybe at most few tens of seconds. And the results need to be stored and properly processed.

Analog TV is nowadays a history, but nearly every analog TV implemented with integrated circuits auto-tuned itself during the beam retrace period, so it was 60 times a second in the USA, and 50 times a second for most of the world. Color TVs did full calibration in steps over 4 frames or 8 fields. Video tape machines did more complex auto-tracking but this was mostly due to limitations of the quite complex mechanical tape transport. In a coin miner the only mechanical variable to tune is the fan speed in RPM.

But anyways, this wasn't some highly technical and patented research. It may have been patented in 197X to 199X, but any patents long expired.

So it isn't like I'm talking about some advanced rocket science or some other secret algorithms. It is all open source now:

https://en.wikipedia.org/wiki/Nelder–Mead_method

and the relevant code is GPL-licensed, exactly the same as -ck's cgminer.

With Spondoolies the situation was different. They intentionally hired an inexperienced developer for full time maintenance work on their code and never paid (or didn't pay in full) the experienced consultants that did the initial work.

I'm really curious what are the reasons for all this technical weirdness in the coin mining business.

Edit: made the link properly clickable.

No one hates u. U dont matter to me. I juz cant stand unethical conduct. It is just too obvious that the one with overly positive marketing and ridiculous non-substantiated claims are of some obvious types of accounts. Anyways, I dont think it matters even if this post gets deleted also. People arent so stupid to believe. That is why the hashrate in those pools and difficulty of btc isnt impacted by this launch. Good luck to u.

If u people have ethics, u would do proper things - reveal yourselves and not use sneaky accounts to make huge claims of greatness,etc.

---

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

im not sure how this isnt "correct". if its anything like the s9 autotune firmware 30 mins of ramping up the freq of each board was more than enough to generate normal opp temps. Anyone that had an s9 when auto tune first came out knows they ran hot during the tuning process which happened everytime the machine rebooted. That was the case with all of mine unless someone can explain how i ended up with the only miners that ran this way i can only assume everyone elses did as well.

also keep in mind that the autotune s9 firmware was changed soon after it was released to lower the tuning because people didnt like waiting an hour for the machine to finally start hashing. I remember hooking everything up and configuring them and swearing i did something wrong (even though its a super basic setup) because no one at bitmain bothered to tell us why the miner was saying "socket connection refused" for almost an hour or why the miner was heating up while it wasnt mining. so me and many others turned them off thinking something was wrong until bitmain told us we had to wait longer before it finally kicks in.

maybe im reading it wrong but ur reply seems pretty hostile...just because u dont agree with the way its done doesnt mean it isnt being done correctly. Or maybe u can test this feature on ur own miner and see if they do in fact heat up while tuning before throwin out ideas of it not being done right?? idk the dude but someone is makin software to better the miners these folks decided to pay....idk if u or ne one else has the right to get that sorta tone ya know??

as for twice a day isnt it fair to say the folks that have these special situations can go to each of the miners and reboot them twice a day so it can retune the miner for the change in weather?? either way do u think folks will be happy with losing hours to let the firmware tune every x hours?? not to mention the wasted power from doing it. im sure that all adds up and eats into ne profits the tuning would provide i would think.