ANTMINER S3+ Discussion and Support Thread

[MoreBloodWine]

I read somewhere that the new firmware update shows a new Freq tab or something, I have just uploaded it but don't see any changes?

Miner Configuration > Advanced Settings.

[1713537628]

1713537628

[1713537628]

1713537628

[1713537628]

1713537628

[1713537628]

1713537628

[Ilan1]

I read somewhere that the new firmware update shows a new Freq tab or something, I have just uploaded it but don't see any changes?

Miner Configuration > Advanced Settings.

Thanks, so if I also use Putty to O/C does this now do anything?

[MoreBloodWine]

I read somewhere that the new firmware update shows a new Freq tab or something, I have just uploaded it but don't see any changes?

Miner Configuration > Advanced Settings.

Thanks, so if I also use Putty to O/C does this now do anything?

If you have the new FW installed you can just forget putty all together.

[Ilan1]

I read somewhere that the new firmware update shows a new Freq tab or something, I have just uploaded it but don't see any changes?

Miner Configuration > Advanced Settings.

Thanks, so if I also use Putty to O/C does this now do anything?

If you have the new FW installed you can just forget putty all together.

Got it, Thanks

[PatMan]

I do wish the BITMAIN would publish the device driver details so that a current version of cgminer could be attained. I know at least 2 releases in 4 had notes about queue improvements and one at least specifically stated it generated work much faster. Also one of the last 3's had an improvement about how work was loaded into devices. I know for sure 3.12 is very old and I am positive that many work generation improvements have been added since then.

Exactly. I don't understand why they don't work more closely with the cgminer devs - using such an old miner for this equipment is false economy. There have been many improvements & optimixations since this cgminer release, not to mention a very important fix for work redirection by a third party.

[buy4crypto]

Wow!  It's even hotter today for my S3s than yesterday.  Ambient room temperature peaked at 37.3C (99.2F).  They have been clocked higher at 237.5M this time around and drawing 408W at the wall but didn't seem to skip a beat at all.  These were the stats during that particular period:

That makes me feel a bit better when mine get up to 43-45c! Almost through the summer here, then a NY winter will keep these ambient down quite nicely from that point on.

[bluecityste]

Hi, if anyone has any coupons left I am looking to buy at 0.025 per coupon.
I think most of them actually run out today, so if anyone has any spares please send me a PM.
Thanks guys,
Stephen

[MoreBloodWine]

I do wish the BITMAIN would publish the device driver details so that a current version of cgminer could be attained. I know at least 2 releases in 4 had notes about queue improvements and one at least specifically stated it generated work much faster. Also one of the last 3's had an improvement about how work was loaded into devices. I know for sure 3.12 is very old and I am positive that many work generation improvements have been added since then.

Exactly. I don't understand why they don't work more closely with the cgminer devs - using such an old miner for this equipment is false economy. There have been many improvements & optimixations since this cgminer release, not to mention a very important fix for work redirection by a third party.

Can we just SSH in to it and compile CGMiner to a newer version ?

[-ck]

I do wish the BITMAIN would publish the device driver details so that a current version of cgminer could be attained. I know at least 2 releases in 4 had notes about queue improvements and one at least specifically stated it generated work much faster. Also one of the last 3's had an improvement about how work was loaded into devices. I know for sure 3.12 is very old and I am positive that many work generation improvements have been added since then.

Exactly. I don't understand why they don't work more closely with the cgminer devs - using such an old miner for this equipment is false economy. There have been many improvements & optimixations since this cgminer release, not to mention a very important fix for work redirection by a third party.

Can we just SSH in to it and compile CGMiner to a newer version ?

One would assume that we have access to their modified code and driver for cgminer which the cgminer license stipulates, however they have ignored our requests for making that code available. I am considering what direction to take now about that license violation.

[J4bberwock]

I am curious has anyone had any RMA issues with Bitmain ?  i have provided them with the order ID and I am still having the hardest time getting RMA information from them.  They have demanded I perform a test with something called a multimeter (which i do not own, i had to go purchase one)  so i could test the chains.  My s3 device pretty much bricked itself with a firmware upgrade.  I cannot access the UI, the ip is not found (never changed from default).  I used an ipsniffer, and still zilch....anyhow, i am curious to know how Bitmain has dealt with RMA requests from others.   Thanks

It took me 2 1/2 weeks of those emails back and fourth. I finally told them I would sell it for parts on Bitcointalk.org and tell everyone about them not honoring warranty. Then and only then did they sent a RMA form. I sent it(controller) off 3 weeks ago and still don't have a replacement. Good luck.

What were your failure symptoms? I had one die today.. going to have to go down this road...

If somebody in Europe/France does have the same trouble, I'd be happy to have a look and try to unbrick it if Bitmain doesn't accept RMA.
Looks like unbricking could be done by serial line programming (easier) or JTAG if serial doesn't work (will have to find the solder points).

Both methods will require soldering and will void your warranty.

Also, what are the tests asked by Bimaintech?

[philipma1957]

S3 sales slowing down or production ramping up?

I don't know about the production, but you can bet the sales are slowing down. Counting current diff increase trend 0.66 BTC per miner may very likely never been mined out of them. Probably expect serious BTC miner price reduction for Batch #7.

Haha, wrong sir!

Interesting. If it's shipped Aug 20th, it means starting mining Aug 23-25th. Type that in any calculator as a starting date with 10% difficulty increase and tell me if the miner will return 0.66 BTC.

10% difficulty increase?!? You wish man.  Next one will be at least 20-25% possibly higher....

30 mil hash came into the network right after last difficulty adjustment.  I agree next difficulty will be in the 20-25% range if not higher.

Also, besides the 30 mil added it looks like more then usual small network hash is getting added.  The 30 mil bumped us up to 169 right after ( 7 mil more was added since that spike )  

I was going to buy some more s3, but looks to me like something new is coming or private placement deal is going on withs oem big manufacturer.  The price staying at .66 btc is not staying in line with bitmain customers at least having a chance to get ROI like in the past.

I for one am holding off on getting any additional s3 units without a coupon ( Not happy bought 12 units and no coupons ) or a significant price reduction.

right now block rate is 8.8 minutes not 10 minutes.  that means 12- 14 %   not close to 20%.  

clock the block rates  if we stay at 8.8 minutes  we need to increase 13.6 % to go back to 10 minutes.

   8.8 x 1.136363 = about 10  .  If the rumor that cex.io added a monster center in iceland is real that was the big jump.  we won't add much from now on.

Care to make a gentlemanly wager?

 No I never make bets.     especially  with christ

my 12-14%  not close to 20% could be correct  after all.

https://bitcoinwisdom.com/bitcoin/difficulty

Bitcoin Difficulty:   19,729,645,941

Estimated Next Difficulty:   22,959,869,961 (+16.37%)

http://bitcoincharts.com

Difficulty   1972964594

Estimated   22005031037  about 11.53%

[mdude77]

Do any reputable sites (non ebay/amazon) have S3s in stock?

M

[black3000gt]

Do any reputable sites (non ebay/amazon) have S3s in stock?

M

Not that I am aware of.

[martijntje]

Anyone have experience with an S3 where chain #1 gets much hotter? I have a 40 degree temperature on chain #1 while chain #2 is only 32 degrees. The whole thing is a lot slower than the other S3 I have.

Have already tried removing all the thermal paste from the chips and applying arctic MX-2, but that didn't change anything. Does it also help to change the paste on the inside? What else is there to try?

I had a similar problem. Mine were off by 4 to 5 degrees. I used the same paste as you, but went one step further and did one of the inside heatsinks,  before I ran out paste.  I have more on the way. The inside heatsinks have hundreds of thermal paste dots over the whole surface, so it was a challenge to try and spread the paste around.
The results were that the blades are now within 2 degrees of each other,  but the unit still struggles to pull 440gh/s, drops over time, and can't be overclocked without lower hash rates and Xs and -s showing up.
After the paste gets here, I will redo the other blade and see if it helps at all.
On a side note, I have also installed heatsinks on the R47s and the little black chips above them, and it did not seem to help at all. At this point I'm a little nervous about them being in there.

Thanks. Seems you have a similar problem as I do, just with a smaller temp diff. My problematic unit also doesn't OC well. I actually ran out of paste after just doing all the chips. Let me know if you see any improvement from repasting the backside of the other blade. Might get new paste and try that as well.

If you say the temps dropped but you didn't see any improvement in speed it might not be the source of the problem actually.

[IYFTech]

I do wish the BITMAIN would publish the device driver details so that a current version of cgminer could be attained. I know at least 2 releases in 4 had notes about queue improvements and one at least specifically stated it generated work much faster. Also one of the last 3's had an improvement about how work was loaded into devices. I know for sure 3.12 is very old and I am positive that many work generation improvements have been added since then.

Exactly. I don't understand why they don't work more closely with the cgminer devs - using such an old miner for this equipment is false economy. There have been many improvements & optimixations since this cgminer release, not to mention a very important fix for work redirection by a third party.

Can we just SSH in to it and compile CGMiner to a newer version ?

One would assume that we have access to their modified code and driver for cgminer which the cgminer license stipulates, however they have ignored our requests for making that code available. I am considering what direction to take now about that license violation.

Indeed, I saw your post requesting access to the code - ignoring it seems a very strange course of action by Bitmain, especially since agreeing & cooperating would not only benefit everyone - but would also stay within the legally binding license terms.......

I hope they realize their error & do the right thing - as well as live up to their claim of supporting the Bitcoin community.

@ckolivas: Is there anything we, the community, can do to help? Maybe if we all started mailing them they will take notice?

[JennaK]

Hashrate - 5min:513.11G / 30min:493.11G / 1hr:487.73G O
Is that alright for 237M?

[Atomar]

I think that's what I have with the default settings. I think it should be more though

[IYFTech]

Hashrate - 5min:513.11G / 30min:493.11G / 1hr:487.73G O
Is that alright for 237M?

Yup 

[JennaK]

I wish the Antminer monitor program worked but for some reason it gives a .net error.

[elemental]

One would assume that we have access to their modified code and driver for cgminer which the cgminer license stipulates, however they have ignored our requests for making that code available. I am considering what direction to take now about that license violation.

I have often wondered about this, it's really not on for a hardware manufacturer to misuse open source software in this way.