Topic: Am I wrong for buying S9's? (Read 235 times)

One speed only and no one wrote after market for it.

I had 4 of them 2 needed the middle board powered off. 2 did not .

All are still running. purchased them in sept 2018 dirt cheap direct from bitmain

The cheapest bulk price in China for S9 when BTC was at sub 3k was $70-80 excluding shipping, anyone who gave an S9 away for free at any given time -- simply wasted at least $70, but you are right about buying it now seems like a bit of a gamble if you don't have free power.



No LPMs.


It's not really a serial number or anything, in the kernel log it was called "chip level" IIRC, and there is a total of 4 possible levels which you get assigned randomly, so if you have more than 4 miners you are guaranteed to be able to mix and match at least 2 miners, some people were lucky enough to get the majority of their S9ks with the same number so they could easily mix and match, but overall, these miners are terrible.

So they run so hot that you physically disable one to keep it cool?, interesting... Doesn't that firmware provide low power modes?

AFAIK the S9K is like the modern miners (S17, etc) where it won't allow you to just swap a hashboard as their firmware records the serial or something.
Well that explains the many failures of the S9k, overheating issues, i see...

Nope s9k are very very very special you are Locked in to bitmain firmware.

Hint disconnect the middle board on your troubled units. the psu part you can leave the data cable attached.

keep all of them very cold.

they run hot.

I had four of them they still work.

But 2 need the middle board not attached.

Buying S9s here does seem like a bit of a gamble, but you never know if it will pay off for you or not.  Not too long ago people were giving S9s away to anyone who would pay shipping.  Now they have a viable resale market.  The concern for you that I would have is that resale will likely be difficult with these, or likely impossible at some point.  So if I were doing the calculations, I would estimate for a higher difficulty, lower price, and no recoup from resales.  If it still makes sense, then go for it.  Remember there's time and effort involved as well.

You have to keep in mind as time goes on the S9 will become unprofitable sooner than later.
I'm only a very small farm miner. An only my opinion here.

With the overall Hashrate at all time highs near 170 Eh/s and new units being cranked out as fast as possible by multiple manufacturers, (only currently slowed by chip an logistics issues right now), Next gen units on the Horizon (s19 XP), and Quite a few large scale Farms up and coming, in all, will eventually send that Hashrate way higher, and sooner than most think in my opinion, and render those S9's unprofitable fast. Total guess, and depends on Bitcoin Price action of course, but Id give the S9 another year of profitability.

Remember Crypto Winter, when the next one happens (and inevitably some sort of one will occur at some point for at least a while I would bet)
Just needs the right combination of Bitcoin Hashrate (difficulty), Bitcoin price drop, and S9's wont be worth running, and selling wont bring much return. I remember when Bitmain was selling S9's new direct for under $100 (before shipping cost) during crypto Winter.

A few S9's for a newbie getting their feet wet may be ok, A lot of learning in setting up a starter farm. But if your going into for the long haul and long profit. Go with the most efficient (new or used) reliable unit(s) you can.

I started in 2017 with 1 S9. Had 13 various machines at one point in 2018.
I have a mix a models (ages) from L3+ to S19 Pro running now (7 units total now)

Buy, run, profit,  HODL, upgrade, repeat.

Good Luck

On that topic...

I'm waiting on a "small" order of 30 S9k miners... 
Got'en for very cheap

Any suggestions on an alternative firmware?
Any way to bypass the signature lock (if they come with the latest-locked Bitmain firmware installed)?
Ideally I'd like to have them pop up in my current HIVE account (HIVE has a "client" firmware for that, but I first have to bypass the sig-lock)

Thanks

+1 to Gabrics' statement.

A major issue for most people running profitability calculations is that they don't consider total cost of electrical: the $/kWh rate listed on your electric bill is probably not the full picture.

And even Gabrics didn't mention another major issue: cooling costs.  One of my S9j boxes throws off 1350W of heat...which is like having a hair dryer running full blast on hot...all the time. Multiply that by even 10 boxes, and you're talking about a TON of heat that you have to dispose of.

...and that's not counting the noise.

Are you also sure that the TOTAL cost is 4.7c not just the power (add transmission, taxes, demand, decommissioning, etc.)?
I mean when you divide the final amount of the bill with used kWh you got the 4.7c?

You are not wrong, but you are missing two crucial points that could change the entire game.

1- At $380 it's probably an S9 (as in S9 without I/J) which is anywhere between 5 to 6 six years old that can die at any given moment.
2- With the efficiency of S9 you are more likely to enter the "mining at a loss" phase A LOT sooner than you do with an S19.

You power rate is indeed below average (since the average 3-5 cents of China is no longer valid) but that doesn't mean you can just buy the least effienct gear and expect to make a profit, my advice would be something in the middle, 50w/th range will be good for your power rate, I recommend whatminer M21s / M20s.

if you have 4.7 cent power and run your s9 at a realistic speed of 13th using about 1050 watts

you are more likely to profit than if you buy 1 s19. it is a bit more complex than your idea and a bit more complex for a few reasons.

I mix my gear to flatten risk. to be very honest i would suggest

l3+
s9
a four card eth rig
and a better btc asic maybe a whatsminer 31s or an avalon 1246

The ways to play a $13000 mining investment are complex.

what matters when you mine

price per kwatt is ace
how many kwatts is king
how well can you cool the space is queen

next hand
price per th ace
efficiency king
durability queen

If you get 1 s19 and it dies you are fully fucked
if you get 36 s9 s some will fail but you can mine with the others.

if you a  btc purist get 26 s9 units set 3800 on the side for power

26 units will burn 18000 kwatts in a month that is around 900 usd in power
you can use the 3800 over 4+ months before power really costs.

a true shit ton of ways to do a 13000 investment

Maybe Whatsminer M31S or Avalon 1246 could also be an option?
Miner Expert seems to have ok-ish prices for those models.
You can see their latest listings at their Telegram channel.
Telegram: https://t.me/minerexpertsales
Website: https://www.minerexpert.com

And ofc you can also ask an offer directly from MicroBT (Whatsminer manufacturer, www.microbt.com ) or Canaan (Avalon manufacturer, www.canaan.io ).
The S9s must be used a lot already, considering it is an old model, so maybe there is some risk for some of the units to break down.

For all my calculations regarding profitability, ROI, power consumption, etc. it seems that the S9, for the prices they're going for now, tend to be more profitable than the newer more expensive S19/S19j's. I know that the power efficiency on the S19 is much better than the S9, but the ROI on the S19 are 2 years + when accounting for estimated difficulty change over time and the current $ per TH. The S9, at current Bitcoin prices, will become unprofitable quicker with the rise of the difficulty than the S19 would; but for the short run am I wrong for investing in S9's than S19's? Am I missing something? My electricity cost is $0.047 USD kWh and as of now I can buy a S19j Pro 104th for $13,780 ($132.5/TH) or for the same amount get 36 16TH (Braiins os) S9's @ $380 per unit ($23.75/TH). Any insight or help is appreciated. Thank you!