ANTMINER S3+ Discussion and Support Thread

[rograz]

Does anyone have the weird problem where normal clocked s3's give 440 GH/s quite well, and overclocked S3's (to 250), give...440 Gh/s?

They do seem to start out faster, with a higher rate initially, but always seem to go back down to this level. I just don't understand why they would do that. Hardware errors are low in all cases. I'm using P2Pool, but I don't think that should have anything to do with it. Does anyone shed some light on the issue?

From my testing they don't seem to show this behavior if you keep them cold, and I mean cold. I stuck a unit that would do 250 for 1-2h and then started falling in performance in front of a portable AC and got the load temps below 25C, it seemed to be just fine at 495GH/s+ under those temps (12h~ of testing) but not really worth it with the AC costs ;P

Or maybe it's just random behavior and it will drop the has hash rate despite the cooling given enough time, someone will figure these out eventually :/

[1714641186]

1714641186

[faetos]

Did anyone receive the refund yet? Mine is still in init status.

[philipma1957]

Does anyone have the weird problem where normal clocked s3's give 440 GH/s quite well, and overclocked S3's (to 250), give...440 Gh/s?

They do seem to start out faster, with a higher rate initially, but always seem to go back down to this level. I just don't understand why they would do that. Hardware errors are low in all cases. I'm using P2Pool, but I don't think that should have anything to do with it. Does anyone shed some light on the issue?

From my testing they don't seem to show this behavior if you keep them cold, and I mean cold. I stuck a unit that would do 250 for 1-2h and then started falling in performance in front of a portable AC and got the load temps below 25C, it seemed to be just fine at 495GH/s+ under those temps (12h~ of testing) but not really worth it with the AC costs ;P

Or maybe it's just random behavior and it will drop the has hash rate despite the cooling given enough time, someone will figure these out eventually :/

It is all about the r47 dc to dc parts  keep them cold you will improve a bit.

heat sinks are cheapest way as they are a one shot price.

the  trick is to figure a way to get some air on them without killing off the rest of the units air flow.  every idea I have thought of that is cheap and easy means putting a few holes into the  case.

right now I have 2 more coming and I am floating a few ideas in my head to test.

[bigbeninlondon]

Was Batch 4 always Aug 7th for ship date?  I could have sworn bitmaintech.com showed Aug 5th yesterday.  Can anyone confirm the shipment date of batch 4?

look closer.
Batch 4-by Aug 2
Bath 5-Aug 7

Yes, I meant Batch 5

[bigbeninlondon]

Was Batch 4 always Aug 7th for ship date?  I could have sworn bitmaintech.com showed Aug 5th yesterday.  Can anyone confirm the shipment date of batch 4?

Almost
It was August 5th for batch 5... then it changed

Typo; I meant Batch 5.  Yea, I thought it changed.

[Sitarow]

Yeah but having your miner queue work from 5minutes ago and gain nothing is also a waste.
Properly setting up a queue/scan-time/expiry can decrease the load of your unit and increase efficiency of the work done.
Proper settings isn't the dark age, it's just something that never is done right There is a reason these options are here for us to optimize with the specific hardware we use!

I wrote the software, I know what I'm talking about. You're barking up the wrong tree with scan time and expiry.

I'm not barking up any tree
You wrote a software and we use the optimal settings for the different hardware and pools we use. Correct?

queue of 4096 is not helping the S3 while a lower queue is, so I'm sorry if you misunderstood my post.

Should we just go put 99999 on all the settings and expect them to function the same?

sir, ghash vs eligius vs p2pool
pls try these pools with the modified queue and scantime...

hope we can see something WOW here

Yeah if you need, just modify your own config, point them at the two pools and conduct your own tests 

My discarded practically disappeared and my load times were dropped by 1/3 1/2... I am happy enough with this. I can't justify why the same overclock is higher now and why the hashrate more stable

Well done. Hope to see more with next firmware updates. Can you see what the power draw is at the wall?

[R4v37]

Yeah but having your miner queue work from 5minutes ago and gain nothing is also a waste.
Properly setting up a queue/scan-time/expiry can decrease the load of your unit and increase efficiency of the work done.
Proper settings isn't the dark age, it's just something that never is done right There is a reason these options are here for us to optimize with the specific hardware we use!

I wrote the software, I know what I'm talking about. You're barking up the wrong tree with scan time and expiry.

I'm not barking up any tree
You wrote a software and we use the optimal settings for the different hardware and pools we use. Correct?

queue of 4096 is not helping the S3 while a lower queue is, so I'm sorry if you misunderstood my post.

Should we just go put 99999 on all the settings and expect them to function the same?

Here is the thing.
Queue is the number of work units to have ready in advance of need. 4k is high. 0 means everytime something runs out of work the CPU right then has to make new work. Having a queue of 1 means it sends work then makes one more.
Expiry is the time in seconds to hold delayed or un submitted work. 1 second means if it can't send immediately it's thrown away.
Scan time is the time that it can work off of given work units. Using stratum you will not need this lower as there is plenty of local work.
P2pool is different some. But for anyone not using it the defaults for cgminer are fine. It doesn't waste tons of bandwidth trying a new unit every second or throwing away work on any network delay. The queue is far above the default of 1.

now i understand what this boss talking bout...
so, its not like we need to modify anything...

[Lowell904]

@lenny_

440GH/s 340W

I'm getting 440GH/s with 378Watt on a CX500 in the US on a 110V circuit, at the wall, taken about 5 min after start-up.

I have 2, each on a CX500. After taking readings I'm thinking it would be more efficient to have them both on a single PSU.

[Lowell904]

Yeah but having your miner queue work from 5minutes ago and gain nothing is also a waste.
Properly setting up a queue/scan-time/expiry can decrease the load of your unit and increase efficiency of the work done.
Proper settings isn't the dark age, it's just something that never is done right There is a reason these options are here for us to optimize with the specific hardware we use!

I wrote the software, I know what I'm talking about. You're barking up the wrong tree with scan time and expiry.

I'm not barking up any tree
You wrote a software and we use the optimal settings for the different hardware and pools we use. Correct?

queue of 4096 is not helping the S3 while a lower queue is, so I'm sorry if you misunderstood my post.

Should we just go put 99999 on all the settings and expect them to function the same?

Here is the thing.
Queue is the number of work units to have ready in advance of need. 4k is high. 0 means everytime something runs out of work the CPU right then has to make new work. Having a queue of 1 means it sends work then makes one more.
Expiry is the time in seconds to hold delayed or un submitted work. 1 second means if it can't send immediately it's thrown away.
Scan time is the time that it can work off of given work units. Using stratum you will not need this lower as there is plenty of local work.
P2pool is different some. But for anyone not using it the defaults for cgminer are fine. It doesn't waste tons of bandwidth trying a new unit every second or throwing away work on any network delay. The queue is far above the default of 1.

now i understand what this boss talking bout...
so, its not like we need to modify anything...

I did queue 0, scan-time 1, and expiry 1 as suggested on a stock S3 and I've been on failover for about 8 hours straight.
Hash rates are about the same both poolside and on the gui. Discards are way down and HW looks down too. Pool has next to no discards in failover but I was getting a lot in load-balance.

Would someone who know what they are talking about please provide some suggested settings, so we can start testing them out?

From above would, queue 1, scan-time 0, and expiry 0, be the way to go?

[Tripjammer]

How big is this batch 5?

[jelin1984]

What exactly is the beeg at s3

[Xian01]

FWIW, After ~10 hours running with --queue 32, slightly overclocked to 225, share difficulty set to 256.

Does not seem to have affected hashing at all, but noticeable CPU load decrease.

Please, what would you suggest we set the queue, scan time and expiry to?

Again, do not touch scan time and expiry.
Giving you generic advice regarding the queue setting is difficult because this is not my driver...I'd be surprised if it really needs to be that high though, and on a low powered system there is rarely anything to gain from anything even in the hundreds, let alone thousands.

Thanks for this. Going to run one of my units with --queue 32 overnight and see how the stats look when I wake up.

[Ilan1]

Still no 7.7% refund  anyone else got there's yet? Can someone from Bitmain please confirm that we will actually receive this today?

[jpchrist]

Still no 7.7% refund  anyone else got there's yet? Can someone from Bitmain please confirm that we will actually receive this today?

Nope still waiting

[ftoole]

I don't see it that way. I see it as part of the risk if you offer paid utilities to someone. The rental agreement should include provisions to restrict or prohibit mining if the "landlord" doesn't want to pay for mining (or grow lamps, for that matter) as part of the electric.

I just recall my dormroom 20 years ago and I (and a roommate) could never have put up with the noise and heat from an S1 for more than 5 minutes.

Hey the s1's are quiet and not too hot. I had several of the 1000watt avolon 200 gh they are loud and hot i think 1 of them was actually worse than a 1th dragon.

[crazyearner]

Has their been any news on the S1 upgrade modules to upgrade with the S3 modules to use or is this something that's going to take longer than expected as am tempted to buy a few new S3 however wanted just the modules so can upgrade my S1.  I was reading that their where meant to be out around the same time as the S3 come out but still seen no updates in regards to upgrade modules to fit the S1 with the S3 modules

[Xian01]

Has their been any news on the S1 upgrade modules to upgrade with the S3 modules to use

No news yet. Curious myself. Have 9 S1's collecting dust I would like to upgrade if possible.

[Biodom]

i have a strangest unit on most recent firmare.
At 212.5 speed is ~429 GH (after 12 hr), almost no HW errors, temp ~40
At 218.75, speed is ~420GH (after same time), almost no HW errors, same temp
At 225, speed is ~420GH after 12 hours (initially bumps to 455, then slowly drifts down), no errors, same temp

What exactly is the limiting factor here?
I am thinking that is is that darned Dc-Dc converter, so would be very eager to learn if someone got an improvement by attaching heat sink, etc.

[crazyearner]

Has their been any news on the S1 upgrade modules to upgrade with the S3 modules to use

No news yet. Curious myself. Have 9 S1's collecting dust I would like to upgrade if possible.

Mine are going to be collecting dust at the end of this month when they get turned off however going into a partnership for mining as opted to pay me for energy costs so going to see how it works outs. other than that if it fails then ill probs end up holding them until theirs news of the upgrade modules and continue to trade on markets.

[JennaK]

Can any Canadians here sell me one?