Topic: Bitmain introduces the Antminer S17 Pro, Antminer S17, and the Antminer T17 - page 23. (Read 11283 times)

I forgot all about this. Not that I was grabbing one just wanted to see the pricing.

Under the noted I think there is a batch 2 on the 10th.

yep - I missed the sale as well... I went to bed at 1230 am Eastern and they were still showing as "TBD" to me, I got up at 630 am and they were sold out... but yes with 2k+ price tags without coupons and with the tariff - it would be a super long ROI

I almost think it would be better if these manufacturers would stop producing products for a while so that people who bought machines could reach ROI before the new ones come out. I think I will just keep an eye out for more T15's because at 1/3rd the price, they are much more likely to make ROI than the S17 ---- though I would love to replace all 20 existing miners with 10 S17's as it would replace my 289 TH/s with 620 TH/s and while it would also replace my approx. 22,585 watts of power usage with approx. 27,900 watts of power usage - the more than double hashrate would compensate for that. Right now, my profit is just around $850 per month - the change over would put my profit around $2,650 per month (based on current difficulty and BTC price of course) to do that upgrade would cost over $25k pushing ROI to almost 1 year at current difficulty/BTC price/profit margin. But that profit will decrease rather soon, especially as now difficulty is going to climb when these hit the market. So, it is hard to justify not upgrading as my existing monthly profit may disappear but to maintain a monthly profit means to push the ROI too far out....

I just need free electricity and then profit will always be easy lol

Not sure how many were in batch 1, but they sold out fast.

Well that didn't last long... Bitmain as usual.



But i don't see any "T17"s, where did you get that info from? I assume "Pro" 53 is the most efficient one?

I slept through the sale.  They were not for sale at 12:50 am and sold out by 6:32 am EST times.
 
The prices  were too high glad I missed it.
2043 to 2366 not counting shipping
I paid 2048 for the 72th m10  that includes shipping.

My S15s will ROI faster than buying any of the new S17 at current prices... Better to wait couple months to buy in my opinion too. Also, many few units were available for sale, so difficulty should not jump heavily.

Same, better wait till 1.5k or less...

Can Almost guarantee that they will be 1.5k by end of this month. I don't think they can make 500$ in a month......unless $BTC goes up....which means that its better to just hold BTC. Idk man, the math isn't adding up right now.

I was ready to pull the trigger on up to 10 PRO's @ up to $1700 each, but $2360? That's $3000 per machine including the tariff to nets > $7 a day at 8 cent power @ current difficulty & BTC price.

Even if I still had the March coupons it would have been a pass. Should have gone to bed.

I got 3 of the 50 TH Pros.  While more than I was hoping for still worth it IMO.

I got mine ordered.

Hmm.........2k + for unit.......... who's willing to bite the bullets first?

OVER THE YEARS  more then one 13th s9  did better then a 14th s9

I have had about  1000  bitmain units from s1 to s15  and they can vary spec a lot.

If you are buying  more then 5 pieces.  I suggest not over spending on the highest specced  ones.

Yeah if they gave you a day in the factory to cherry pick  the 5 best why not.  But I have seen higher ranked not that good.

I had 1 s7 that was unreal  it clocked  like 50mhz better then any of my others and it was super good on power like 235 watts a th  which was super good for that unit.  It also never tossed hw's 100 days at a time without one.

I honestly feel it's like inno did for the a9.  Ones that were made with better chips that hit 53, they will see as the 53th normal pro, ones that hit 50, they will sell as such for slightly less money.  Both are same efficiency in low and turbo power mode and both do same hashrate in those two modes.  That being said, I honestly see like no difference.  If the 53th is like.. $100 more, sure, if it's like $500 more, lol then no.  It sounds to me like the same miner just maybe more precise chips in the one with 53th is all.

I'm happy with my S15 I won't buy the latest S17 unless it's cheap ( unlikely ).
It just pack more hash and watts but to benefit fully of the new efficiency,
you need to run in low-power mode, cutting precious hash rate for ROI.
In full power, they do a little bit better than the S15 efficiency
but did not justify another upgrade for me.

I sold off all my miners about a year ago but news of this release has me interesting in possibly dipping my toes in again. One question I have though, why did they break this release up into 4x different models? On their site it shows 2x different Pro models (53THs and 50THs) and 2x different standard models (56THs and 53THs).

my set up should be able to handle these as each breaker is 30 amps and I am currently running 2 miners off of each breaker - this would put it as one miner per breaker -  I could run 10 max

I want to buy them to get the early jump on the hash rate before difficulty climbs thru the roof but I am wanting to also wait for coupons to help offset the 25% tariff - that makes it even worse to swallow a 2k price tag.... The last time I paid 2k for a miner, it took over a year to make ROI on that miner - I would like to not go a year to make ROI again.

so - even though I would love to get 10 of these and replace all my S9's - the price is really going to be the deciding factor. I may just wait for coupons and buy more T15's and make ROI hopefully faster.

All out direct dc panel to miners have been a fail.  Can't stay  stable say 12 volts (L3+) with out a ton of batteries and caps  that then makes it too costly.  every dc to dc psu  wants very stable dc in.

Our off grid setups have not worked well so far.

We wanted a doghouse with 4x 320 watt panels to run an L3+  lots of time and money tossed at products that don't work well.

All off topic though.  

Yeah a few s17's will get ordered.

Maybe only 1 at a time as 2000 + 500 trump tax =  2500

Say the pro doing  62th at 2800 watts  62th is  370 a month usd  in btc I get ½ or 185 a month so 2500 /185= 13.5 month for payoff

if gear is 2300 I could be looking at 2875/185 = 15.5 months.

while the m20 doing 72th  earns 430 usd a month in btc I get ½ or 215  I paid 2048 + 560 = 2608/ 215 = 12 month payoff

so m20  is just as good as a s17.

Your post mentioned a shipping date of May 20th: https://bitcointalksearch.org/topic/whatsminer-m20-is-announced-and-available-for-pre-order-5120959
Assuming May 20th, it's just a month of extra mining time, worth ~$200-250 USD. They may be able to charge a small premium for the 12%~ more efficient pro version, but the standard S17 will not sell well if it costs more than the M20s + 1 month mining time.

I wish I had lower electricity costs, 7.5 cents just isn't viable for btc mining.

Yes there is a chance they may cherrypick those units (and your pending M20 as well), so there is no choice but to accept an early review, and then later a second review when "production" units are available. So, an early review, and then again another from a following batch etc. Can be from different reviewers, don't take all the burden yourself

I'm guessing the S17 will be 40~45 watt a th. Since you have a voltage regulator you can measure them 240v, 220v and 208v for science.
If i were to guess, Bitmain may be publishing stats based on the chips themselves (20v DC), or the whole unit plugged to Chinese mains, which is 230v@50hz.

All that current transforming causes some loses. One thing to always keep in mind for you people using renewable, is figure a way to avoid inverting in the first place. I mean, if your panels make 24v and the Asic boards take 20v, shouldn't the whole dc-ac, ac-dc be skipped altogether? If you can ever pull that (skip the PSUs, direct feed the units) you are going to be more efficient than everybody else.

To begin with DC tends to lower its voltage on its own, use a long enough wire and it might already be down to 20v by the time it reaches the unit... Else you could figure out DC to DC psus, they exist for regular PCs and servers (especially useful if your input DC is too different, ie. 48v or 12v, or mostly to buffer voltage variations from the input (such as when you turn on/off another unit) and ensure steady 20v to the output.

The top numbers for power draw that I marked in red use as much as 3300 watts so 20 amp is needed to be safe.

Note that is pro model in a hot room on turbo 3371 watts.

Lastly 200 to 240 volts. Means my garage is out as it can drift up to 245 volts.

My spot in Clifton NJ  is better 226 to 229 watts.