Topic: Bitmain Introduces the S17+ and T17+ - page 2. (Read 3389 times)

News about high failure rates doesn't surprise me much. I've been doing some learning on S17/T17 hardware a bit lately.

They have substantially improved airflow and heatsinking in ways which should have been obvious to them in 2015. However, depending on the model of miner, they're pulling anywhere from 14 to 21 watts PER ASIC, from the same 8x8mm package size and footprint they've been using since the S9. The package is improved with a huge exposed die for better top-cooling junction/sink connection but that's still roughly double the power per chip compared to previous miners.

I totally agree with this!

Also, when talking about luck... Imagine this (my story):

Getting S17 Pro (first batch) and all 3 hashboard end up failing after 2-3 months. Not too bad, I can still RMA it as it is in warranty, won't cost me a lot (excluding time and shipping costs). I get new S17 Pro (after some problems with my government) and it ends up (after 1-2 months) with 2 dead hashboards and one popping up from time to time. Now I am a bit angry... It's same problem all over again, but now I am ready! I can avoid problems with my government and I hope price of fixing my gear will not be too high...

Lucky me it cost me ~$150 in shipping and $55 for repair cost. BUT, due to stupidity of my government and problems with pandemic, I had to pay VAT (20%) for the third time FOR SAME MACHINE, and lost about 5 months of mining only in waiting for shipment and government to decide wtf they want.

And that is the first unit ever I had problems with. I ordered:
- 20-30 S9s and all of them are still working (not for me however)
- 1 S15 still working (not for me)
- 2 T17+ still working (not for me)
- 5 T17e still working (not for me)
- 5 S17 Pro (1 had the problem I explained above and other 4 are working, but not for me)
- S17+ and S17e (still working and in my possession)
- 6 Z11 (still working but only 1 is in my possession).

Also, of all the PSUs I ordered, none had failed so far (ordered between 30-50 APW3+ and APW7)

So, that is quite a low failure rate for me, but it cost me a lot anyway.

No problem.

Apologies Mike.... I had to look back a few pages. I got you confused with that other guy that was posting all his negative opinions.

(OPP - Old People Problems
Check forum once every few days, see that same guy bashing.
See another negative post, dont look at user name before getting snippy)

Ok, let me break it down for you since you either have a comprehension issue or you are too lazy and have not read the article, and for the sake of simplicity, I like to believe the latter is the case.

The article contains words from Jihan Wu himself saying


Stryfe, sorry my previous reply was not for you, it was for PopoJeff, by all means, it's a pleasure to read your experience.

I don't manufacture or get paid to shill other so my opinion is also 100% unbiased. I simply put my experience down. I am sure a large farm would yield different results as they have more machines and thus more chances of failure.

but is there an issue with people sharing their experience?

Then why do you keep pushing your opinion, if you're going to add a little disclaimer that you dont have enough data to form an opinion?

We got it, you read an article where Chinese miners report 30% failure rate.

Got it. Thanks.

I read articles reporting btc will go to 100k.
I read articles reporting btc will drop to 2k.

As with any piece of information on the internet, I trust it as much as I trust my government.

Luck and quantity matters, you may buy 2 gears and you get lucky with them, that's why only studies with a large number of samples can be taken seriously. also, the time factor is very important, these failures don't really show up in the first month or two, it takes time, the reason why I tend to trust the inputs coming from Chinese farms more than your words or even actually more than my experience is the fact that they have a lot more miners and have run them a lot longer.

Since I don't manufacture mining gears and I don't get paid to shill other mining equipment, my view is 100% unbiased, but by all means, go ahead buy whatever gears you want.



The article explicitly talks about the 17 series, I have vouched for the robustness of S9, S9i, S9j (not the s9K) a few times, I am not sure about the 17 series, I have neither bought enough of them nor have I run them long enough to judge with accuracy.

Oh, I forgot to add two problems Ive had.

I overheated an s17e with bad air flow. It dropped a board and wouldnt detect one board until i did about 10 restarts. Then its been fine.

I also had an S9i last summer that did something (forget exactly what).  But I disconnected the PSU, blew into the connectors like and old Atari cartridge, plugged it back in, and it was fine.

So theres that

I have had a lot of bad:

s9's
s7's
s1  first miner I ever purchased from them came with a loose cap and would not boot.

I have had 4 T17's, 20 S9's (4 vanilla 8 i's and 8 j's), 2 S15's, 1 L3, 3L3+'s, 4 L3++'s, 1 D3 and even though that is not a lot of miners (35) I have only ever sent one back and it was a vanilla S9.

edit: I forgot to add one more S9 that I picked up from a garage sale for $20 - it had burnt connectors on the hashboard (the little plastic part has those dividers in between the pins - those were melted) and a bad psu (the wires had somehow melted and shorted out, I believe that is what lead to the burnt/melted connectors on the hash boards) but the miner and psu still worked once I swapped the psu wires and the burnt connectors.

15 s17 pro all good

  1 s17.           good

  2 t17            good

 2 t17e           good

  2 t17+.       5 of 6 boards good.

  1 t15            good

  1 s15             good

 4 s9k             good

purchased all of the above new from Bitmain feb 2019 to April 2020 one exception the  [s15 was used]

that is a loss of a 20th board on the t17+ which kind of wants to work

all are spec the then 2x t17e which are power hungry.

I would consider that to be really good.

about 1255 out of 1275 hash works. and very close to power specs.

My personal experience:  zero failure rate.   So, I'll stick with what works for me. But thanks for your input.

I found a very interesting read > https://news.8btc.com/bitmains-antminer-17-models-drawing-criticism-from-miners-for-high-failure-rate

Basically Chinese miners are reporting up to 30% failure rate on these 17 series, and Jihan Wu


This may explain how the S19 isn't selling fast and almost always in stock, who would risk nearly $3000 for a gear that may potentially fail at a rate of 30%?


My personal experience so far.  2*S17+ > PSU/Hashboard , T17 > Temp sensor problem and S17 pro > So far okay.

So for anyone who is still buying these gears, be prepared for a 30% RMA rate or higher.

ahh well maybe mine will be here soon. we both had the 27th originally for a date.

My T17+ just arrived it is running okay.

Everything seems to be moved to June except T17+ and spot orders.

Also, S17+ has been lowered from $1320 to $1232. While T17+ still has it's raised price of $860.

As noted by minefarmbuy, S19 has been restocked.

However, even though Serbia is slowly being opened and lowering quarantine limits, Bitmain still does not allow orders for me.

Looks like more inventory on the s19 models as well. Batch time changes on everthing.

Yeah I had a apr 27 date it has since shifted to may 1

it is in USA it left Cincinnati and should arrive in Newark later today. In the pre cover-19 days. I would expect it on weds the 29th.  They still list it for May 1 I will take that date.

My order just flipped from showing the 27th for delivery to showing "pending" with no further info. My guess is next Monday.