ANTMINER S7 is available at bitmaintech.com with 4.86TH/s, 0.25J/GH

[Biodom]

Here is the difference between stock firmware and oct 23 600 firmware running 15 s7's against nicehash.
If you look at the start on the left for about an inch that is the stock firmware running.
After that all oct 23 firmware. The downward spikes are rebooting on my end for naming configuration.
Two more s7's from oct 12 order will arrive tomorrow.

Thanks for the image tip dmwardjr.


https://i.imgur.com/Ei7K1vH.jpg

thanks
not familiar with their interface.
what's green and what's blue? are we comparing first inch of green with the rest of the green or first inch of blue with the rest of blue?

First inch of blue histogram with the rest.

sorry, but first inch of blue is essentially at zero-so, it did not work with previous software, but works with current or you are talking about a second inch compared with the rest. i see some increase there (maybe 5% or so).

[1714683117]

1714683117

[1714683117]

1714683117

[wolfen]

Here is the difference between stock firmware and oct 23 600 firmware running 15 s7's against nicehash.
If you look at the start on the left for about an inch that is the stock firmware running.
After that all oct 23 firmware. The downward spikes are rebooting on my end for naming configuration.
Two more s7's from oct 12 order will arrive tomorrow.

Thanks for the image tip dmwardjr.

https://i.imgur.com/Ei7K1vH.jpg

Very interesting. I assume everything was set the same, Frequency extranonce etc? Would be instructive if Bitmain would let us know what the differences were?

Rich

Everything identical, extranonce of course.
Would be nice for a blurb from bitmain but "works better" comes to mind.

[AriesIV10]

I just received a shipping notice from UPS for delivery tomorrow for an order from 10/18; although, it is still showing Unshipped in Bitmain.      I still have more orders on 10/24 and 10/25.

was it b2 or b3? it would be helpful to know.
my 10/16 b2 order is still nowhere to be seen.

Why did they make batch 2 available anyway if they don't deliver it at least on par with b3, without even considering that one might expect b2 to be delivered BEFORE b3 as the common logic suggests.

The order was for Batch 3.  

[wolfen]

Here is the difference between stock firmware and oct 23 600 firmware running 15 s7's against nicehash.
If you look at the start on the left for about an inch that is the stock firmware running.
After that all oct 23 firmware. The downward spikes are rebooting on my end for naming configuration.
Two more s7's from oct 12 order will arrive tomorrow.

Thanks for the image tip dmwardjr.


https://i.imgur.com/Ei7K1vH.jpg

thanks
not familiar with their interface.
what's green and what's blue? are we comparing first inch of green with the rest of the green or first inch of blue with the rest of blue?

First inch of blue histogram with the rest.

sorry, but first inch of blue is essentially at zero.

First inch of blue mass where the new miners are running.
First inch is just below 75TH, after that pretty much right on it.

[pj40]

Well if I can run one of these bad boys off of 2 HX850's, I am going to have to sell my last S5 and S1-S3+ upgrade (OC'd and hashing at 500 which is sweet and reliable) just to get one out of the next batch so my 4 HX850's can be used.  Guess I'll have to test it first to see if I can reliably do that.  I worry about pumping 800 Watts through a PSU that's rated 850 Watts.

It would probably work, but the S7 I'm running of 2x PSU's are a lot more unstable than the ones I run on 1 PSU. 100 GH/s lower as well, and can't handle overclock at all.
100 GH/s more on the ones I run of Bitmain's PSU and an EVGA 1600W G2.

Multi-meter readings show a 0.1V difference on the dual PSU setup, compared to 1x PSU. So the problem might actually be that the PSU's themselves. Running 2x boards on XFX1250 and 1x board + controller of a cheap ass Corsair 650W.

Hmm, well I'll definitely have to do some testing with my 4 PSU's on my B3 that's coming tomorrow.  I was originally planning running 3 PSU's with it, 1 for each board +controller.  I know they can handle that.  But if I can squeeze out enough juice to power a second S7, I'd rather do that than my S5+S3 combined mostly for the energy efficiency when it comes down to it.  If I have to consider a new PSU, I'll have to evaluate that cost with the next S7 batch to determine if ROI is possible.

[Xian01]

I worry about pumping 800 Watts through a PSU that's rated 850 Watts.

In theory, you should have no issues with pumping 800W sustained out of an 850W power supply. 50W is a decent margin.

[Tupsu]

I worry about pumping 800 Watts through a PSU that's rated 850 Watts.

In theory, you should have no issues with pumping 800W sustained out of an 850W power supply. 50W is a decent margin.

In practice, Corsair RM 850 for example.

[pj40]

I worry about pumping 800 Watts through a PSU that's rated 850 Watts.

In theory, you should have no issues with pumping 800W sustained out of an 850W power supply. 50W is a decent margin.

In practice, Corsair RM 850 for example.

So an HX850 should be fine, the only difference I see is that the RM850 is fully modular while the HX850 is partially modular.

[wolfen]

Here is the difference between stock firmware and oct 23 600 firmware running 15 s7's against nicehash.
If you look at the start on the left for about an inch that is the stock firmware running.
After that all oct 23 firmware. The downward spikes are rebooting on my end for naming configuration.
Two more s7's from oct 12 order will arrive tomorrow.

Thanks for the image tip dmwardjr.


https://i.imgur.com/Ei7K1vH.jpg

thanks
not familiar with their interface.
what's green and what's blue? are we comparing first inch of green with the rest of the green or first inch of blue with the rest of blue?

First inch of blue histogram with the rest.

sorry, but first inch of blue is essentially at zero-so, it did not work with previous software, but works with current or you are talking about a second inch compared with the rest. i see some increase there (maybe 5% or so).

Yes the second inch, 5%, thank you for clarifying that for me sir.
I watch these graphs all day every day so I took way too much for granted on the explanation.

[Xian01]

So an HX850 should be fine (?)

In theory. The initial RM750's and RM850's were a bit notorious for running into thermal issues and potentially causing all sorts of havok. I believe Corsair has since replaced the inferior components they were using in the RM series.

[wolfen]

I worry about pumping 800 Watts through a PSU that's rated 850 Watts.

In theory, you should have no issues with pumping 800W sustained out of an 850W power supply. 50W is a decent margin.

In practice, Corsair RM 850 for example.

So an HX850 should be fine, the only difference I see is that the RM850 is fully modular while the HX850 is partially modular.

80% is max I would do for safety. 60% is safer.
I'm running 3 corsair rm1000 watt psu's to power two s7's, that is at 80% for two boards on s7. At 80% it keeps the fan running slowly to cool.
When you drop $2000 on a rig, don't skimp on psu.
Even a corsair AX1200 is good. Good price on ebay.

[pj40]

I worry about pumping 800 Watts through a PSU that's rated 850 Watts.

In theory, you should have no issues with pumping 800W sustained out of an 850W power supply. 50W is a decent margin.

In practice, Corsair RM 850 for example.

So an HX850 should be fine, the only difference I see is that the RM850 is fully modular while the HX850 is partially modular.

80% is max I would do for safety. 60% is safer.
I'm running 3 corsair rm1000 watt psu's to power two s7's, that is at 80% for two boards on s7. At 80% it keeps the fan running slowly to cool.
When you drop $2000 on a rig, don't skimp on psu.
Even a corsair AX1200 is good. Good price on ebay.

But an AX1200 can't power a single S7 since they use 1210W.

[wolfen]

I worry about pumping 800 Watts through a PSU that's rated 850 Watts.

In theory, you should have no issues with pumping 800W sustained out of an 850W power supply. 50W is a decent margin.

In practice, Corsair RM 850 for example.

So an HX850 should be fine, the only difference I see is that the RM850 is fully modular while the HX850 is partially modular.

80% is max I would do for safety. 60% is safer.
I'm running 3 corsair rm1000 watt psu's to power two s7's, that is at 80% for two boards on s7. At 80% it keeps the fan running slowly to cool.
When you drop $2000 on a rig, don't skimp on psu.
Even a corsair AX1200 is good. Good price on ebay.

But an AX1200 can't power a single S7 since they use 1210W.

Correct, that would be for two boards.
Here are corsair rm1000 used I just googled. $115 shipped. Half of mine are used.
https://www.sierracomponent.com/product/components-peripherals/power-supplies/cp-9020062-na-corsair-rm-1000-watt-power-supply-801940120595-used/

[pj40]

I worry about pumping 800 Watts through a PSU that's rated 850 Watts.

In theory, you should have no issues with pumping 800W sustained out of an 850W power supply. 50W is a decent margin.

In practice, Corsair RM 850 for example.

So an HX850 should be fine, the only difference I see is that the RM850 is fully modular while the HX850 is partially modular.

80% is max I would do for safety. 60% is safer.
I'm running 3 corsair rm1000 watt psu's to power two s7's, that is at 80% for two boards on s7. At 80% it keeps the fan running slowly to cool.
When you drop $2000 on a rig, don't skimp on psu.
Even a corsair AX1200 is good. Good price on ebay.

But an AX1200 can't power a single S7 since they use 1210W.

Correct, that would be for two boards.
Here are corsair rm1000 used I just googled. $115 shipped. Half of mine are used.
https://www.sierracomponent.com/product/components-peripherals/power-supplies/cp-9020062-na-corsair-rm-1000-watt-power-supply-801940120595-used/

ah, ok I misunderstood.  So I'll just stick with my plan of 3 HX850's for my S7, and have to deal with buying another psu if I manage to snag another S7.

[notlist3d]

I worry about pumping 800 Watts through a PSU that's rated 850 Watts.

In theory, you should have no issues with pumping 800W sustained out of an 850W power supply. 50W is a decent margin.

In practice, Corsair RM 850 for example.

So an HX850 should be fine, the only difference I see is that the RM850 is fully modular while the HX850 is partially modular.

RM or HX are both pretty high quality PSU's as far as Corsair's models.   I like the RM series just as I have used so many RM1000's but that is personal preference.

After looking at PSU's lately I think I would look into server PSU's.  A lot of them really are quite a bit cheaper.  It's honestly kinda nice after getting use to using them.

[aurel57]

Hello to you all.
Received my ANT7 Batch 1 yesterday.
Also received the PSU from Bitmain.

Yesterday i pluged everything, configured and "voilá"... it was mining...

Everything OK by 2 hours... after that the unit started to beep.. i needed to poert it off.
Connected again... and once again it was mining.

After 2 hours the some problem... the mining decrese to almost 0, the unit started gain to beeb.
Power off and power down again
The unit beeps again and again... works for 3 to 4 minuts until it stops mining

From 21PM yesterday the unit is producing ZERO... its beeping almost shure (i'm in home, and the unit in in my office).
The ANT7 is connected to a swhitch and then the switch is connected to a router.

Im a newbie, and i thought this was power on and let go... i see its more complicated for a 2500€ machine.
Is there any troubleshotting guide?
Help is needed.

Did you configure your pool information on the miner? As it will mine but to Bitmain account. If you did I think I would check the miner status page and watch your temps see what they are. Then if that looks good I would try and use a volt meter and check your voltage from the PSU under load. Watch it the whole 3-4 mins and see if the voltage drops off as the miner drops off.

I would like to see a screen shot of the miners status page.

[cryptichermit]

OK, believe what you will.  I do know several "licensed" electricians and had one confirm with me today, after these exchanges, that it is Per Phase on the Main - not total.
So each phase can be loaded with 80% load and the main will not trip as they are out of phase and not an addition of current.

Here is a list of miners and wattage that I am currently using - all at 120V.  They alone are exceeding 200A in power usage -

13 x s7's = 15730 WATTS
8 x s5's = 4720 WATTS
1 x S4 = 1400 WATTS
3 x A2 @ 88 = 2250 WATTS
4 x A2 @ 110 = 4000 WATTS

TOTAL WATTS = 28100 @ 120v = 234A

The only 240V circuits being used are 1) Stove - 2) Water Heater.
Even with the miner loads, I'm still able to use all sorts of other electricity - IE this computer - a couple 1000 Watt gaming machines, a 200 Watt Laser Cutter, air compressors, electric drills, radios, garbage disposal, hair dryers (when gf needs),  and charge my Chevy Volt.

You are saying this is impossible, unless defective - seems to be working.

Don't forget.  You know we have 200 amps on each pole but how many volts is on each pole?  The answer:  120 Volts

Yes, exactly.  So where do you think the 240V comes from?  Combining the 120V across the 2 poles.  Essentially when you use a 240V breaker and have a 5A load on it, you are pulling 5A across both pole/legs at 120V.
So in order to achieve 200 amps at 240V you'd be pulling 200amps across both legs of 120V.  This alone, seems to tell me that yes you can use 200A of 120V on each leg.  

Yes Stove/Oven, Water Heater, and Dryer (which I don't use) all utilize both legs with 2 pole breaker.  Who needs a dryer with this many miners?  Just put up some clothes racks in the living room when its time to dry.

I see now you say you don't ever cook, you don't use the dryer and you always take cold showers.  [I'm commenting on your post while I read it.]  I imagine you turn the A/C on during the summer.  Then we have the fridge, the compressor you mentioned, the 1,000 watt gaming you do, the lights, the television(s), your exhaust fan for heat; any other box fans you may use for cooling, your dish washer, etc...

Cheers,

LOL,  I never said I don't ever cook or that I take cold showers - to the contrary I just made some spinach egg scramble as well as some sausages for breakfast (While heating my coffee in Microwave) and am about to take a nice HOT shower.
Clothes, yes - I said I don't use dryer - I air dry them.  A/C - I simply don't use either.  In fact it's disconnected from the main and have 4 x 20 amp 120 circuits installed in its place for miners.  The Fridge operates on a 120v circuit.  I do next to zero gaming.  My son does the gaming, but I do have a dual xeon rendering rig that gets kicked on from time to time that draws 1000 watts when rendering, or mining.

[dmwardjr]

LOL,  I never said I don't ever cook or that I take cold showers - to the contrary I just made some spinach egg scramble as well as some sausages for breakfast (While heating my coffee in Microwave) and am about to take a nice HOT shower.
Clothes, yes - I said I don't use dryer - I air dry them.  A/C - I simply don't use either.  In fact it's disconnected from the main and have 4 x 20 amp 120 circuits installed in its place for miners.  The Fridge operates on a 120v circuit.  I do next to zero gaming.  My son does the gaming, but I do have a dual xeon rendering rig that gets kicked on from time to time that draws 1000 watts when rendering, or mining.

Thanks for clearing it up.  The confusion came when you said you were using 32,000 Watts of rigs and everything else in the house worked fine without tripping the breaker.  The very fact you have the AC disconnected will have everything else work fine with the rigs running.  The AC is the biggest burner of power after the rigs.

[dmwardjr]

dmwardjr-------I have to say that you are a very good person!!!!   You have helped me in the past and are continuing to help other to the best of your ability.  Just from me, Thank You!

Thank you,

My apologies for the delay.  I've been sleeping.  I was up until 5:25am this morning.

I do my best.  I still have a lot to learn with mining, nodes, the blockchain, code, networking, etc...  There are a lot of fine gentlemen and ladies on here with loads of experience and knowledge to learn from.  I feel privileged to be in the company of most all of them on here.

Have a good day, AriesIV10.

David

[edonkey]

It's been a while since I've owned a Bitmain product. I forgot how nice and detailed the stats are.

I've been trying to tune my rigs for optimal performance. With my S3s I used a technique of computing the accepted difficulty per second over a reasonably long time (like a day). Then I'd tweak until there was no more improvement. The idea is that while the other local stats are interesting, the accepted shares is most important.

However I'm finding that this is not working with my current S7 and pool set up. Even for the same settings on the same S7 I'm seeing fluctuations of a 1 or 2 percent in the accepted difficulty rate over a day or so span.

Maybe it's the pool being inconsistent? The rigs have mostly been pointed at F2Pool.

Or is it that the local difficulty is set so high (16.4k) that I'm actually seeing local variance? That seems unlikely.

I guess that if the accepted difficulty rate is not a reasonable way for me to measure tweak results, I have to fall back to raw local hashrate and hardware error rate. If so, what HW error percentage rate are you guys trying to target?

Alternatively if this is pool related, I can switch pools. Currently I've been trying the trick of nicehash as the first pool with a high profitability value, and F2Pool as second. This results with mostly mining on F2Pool and occasionally redirecting at nicehash when it's high enough (like it was yesterday for an hour or so).

Maybe I'll go back to slush. I've mined there a lot since they went beta with their new interface. I only left because I wasn't convinced that they had solved their recent orphan blocks problem (which was related to their pool implementation and the malleability attacks). Also, due to its scoring algorithm you really can't bounce between pools so the nicehash trick is out when mining at slush.