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

[Prelude]

Hi Guys - I need help on upgrading FW on the S7s to the latest ones.
The Bitmain support page lists a number of FW for different S7 batches and I need 2nd pair of eyes to double check which FW goes to which S7.

My current list as follows:

1. 4.73TH/v.1.32 (700Mhz) - Batch 8
2. 4.46TH/v.2.0 (600Mhz) - Batch ?
3. 4.66TH/v.1.7 (600Mhz) - Batch ?
4. 4.66TH/v.1.7 (600Mhz) - Batch ?
5. 4.73TH/v1.31 (700Mhz) - Batch 8

Appreciate any help.

Regards
TMT

I second the opinion ^^^^: why upgrade? It's fine w/o.

Yes, I kind off agree with that rationale however during failover situation with kano site this week, some of the S7s will continue beeping even when kano is up again. The users with Dec FW applied didnt seems to be affected by this. And that is why, it's probably best to get things up to date before New Year.

That's far from a scientific conclusion, but the risk is yours to take.

[1715116880]

1715116880

[VirosaGITS]

Hi Guys - I need help on upgrading FW on the S7s to the latest ones.
The Bitmain support page lists a number of FW for different S7 batches and I need 2nd pair of eyes to double check which FW goes to which S7.

My current list as follows:

1. 4.73TH/v.1.32 (700Mhz) - Batch 8
2. 4.46TH/v.2.0 (600Mhz) - Batch ?
3. 4.66TH/v.1.7 (600Mhz) - Batch ?
4. 4.66TH/v.1.7 (600Mhz) - Batch ?
5. 4.73TH/v1.31 (700Mhz) - Batch 8

Appreciate any help.

Regards
TMT

I second the opinion ^^^^: why upgrade? It's fine w/o.

Yes, I kind off agree with that rationale however during failover situation with kano site this week, some of the S7s will continue beeping even when kano is up again. The users with Dec FW applied didnt seems to be affected by this. And that is why, it's probably best to get things up to date before New Year.

That's far from a scientific conclusion, but the risk is yours to take.

Turning the S7 off and then on again would probably work too. Its rare that Kano gets down/DDoS and i have not seen this happen on simple stratum restart. Anyhow, good luck, would suck if you lost a controller on an upgrade, because afaik, very little "updating" has been actually done by Bitmain that actually fix anything.

[Tupsu]

Hi Guys - I need help on upgrading FW on the S7s to the latest ones.
The Bitmain support page lists a number of FW for different S7 batches and I need 2nd pair of eyes to double check which FW goes to which S7.

My current list as follows:

1. 4.73TH/v.1.32 (700Mhz) - Batch 8
2. 4.46TH/v.2.0 (600Mhz) - Batch ?
3. 4.66TH/v.1.7 (600Mhz) - Batch ?
4. 4.66TH/v.1.7 (600Mhz) - Batch ?
5. 4.73TH/v1.31 (700Mhz) - Batch 8

Appreciate any help.

Regards
TMT

For 2,3,4 use SD-S7-20151023-575M.tar.gz   (support 6 blade in one controller) or
this SD-S7-20151211-575.tar.gz  (3 blade only)

For 1 and 5 use SD-S7-20151211-700.tar.gz

[fr4nkthetank]

Hi Guys - I need help on upgrading FW on the S7s to the latest ones.
The Bitmain support page lists a number of FW for different S7 batches and I need 2nd pair of eyes to double check which FW goes to which S7.

My current list as follows:

1. 4.73TH/v.1.32 (700Mhz) - Batch 8
2. 4.46TH/v.2.0 (600Mhz) - Batch ?
3. 4.66TH/v.1.7 (600Mhz) - Batch ?
4. 4.66TH/v.1.7 (600Mhz) - Batch ?
5. 4.73TH/v1.31 (700Mhz) - Batch 8

Appreciate any help.

Regards
TMT

I second the opinion ^^^^: why upgrade? It's fine w/o.

Yes, I kind off agree with that rationale however during failover situation with kano site this week, some of the S7s will continue beeping even when kano is up again. The users with Dec FW applied didnt seems to be affected by this. And that is why, it's probably best to get things up to date before New Year.

Hmmmmmmmm.  Did you have backup pools ?  I put (i know i know, but what can you do) antpool with pps and f2pool with pps as backup when Kano.is is down.  You only mine a little there once in a while so its not a big deal, but pps is good for that reason.  my batch 8 switched to backup pool, and then switched back to kano no problem.

[fr4nkthetank]

Oh and why arent you people mining over at Kano pool more ?  It takes 100 hours or so to "ramp" up your hash power (it isnt lost, just...much delayed) so you dont pool hop, etc.  put pps pool on backup.  0.9% fees, but they pay transaction fees, which are around 0.9% on average

I mean, check this luck out:

                                      tx with fees                                                  total luck
Last 5 Blocks   2.5days   100.7%   59.39%   59.39%   0.1796   168.38%
Last 10 Blocks   7.6days   101.0%   92.96%   92.96%   0.4515   107.57%
Last 25 Blocks   20.1days   101.0%   95.64%   96.45%   0.4551   103.68%

[SLEI]

I got my batch 8 yesterday which was shipped direct from stock and it is totally different (and 2BTC cheaper) than my 23.11. 2015 shipped batch 8.
Early shipped is only using 1210w from wall and hashing about 4600 with BM power.
This new one is using whopping 200W more with BM power and 4730 Gh/s, same amount of power as batch 7 5300Ghs!!
Batch 8 miners are in same room side by side and running 700Mhz.

I have cut half (6) of the static finns from the exhaust fan of the early bird and inlet fan is running 3600rpm out 3480.
Later batch 8 has different exhaust fan, only 9 static fins but bigger hub area, inlet fan has to run 4200rpm to keep it cool ex 4080, error rate is only 0.0009%.  I conficured the latest miner in colder room for same specs as early bird and didn't notice the different fan and 80C protection had kicked in after 2 hours.

is freq locked at 700 or can you lower it a bit?

say 687.5

 mining the extra 130gh at a price of  200watts  is losing money as you are paying more then 1.5 watts a gh

Watts to Ghs are very linear at my S7s with APW3-12-1600 PSU, early bird has 0,26 J/Ghs and I have tested it down to 400Mhz and this new one 0,3 J/Ghs:  675Mhz 1375W 4550Gh/s; 700Mhz 1410 4730 Gh/s.

If the electronics at my two batch 8 are same, firmware and/or core voltages have to bee different.

I got xxxs on some chips at boot on 700Mhz at Early bird when adjusting fans, error rate is 0,021% and no point to go over 700Mhz.
Late birds 0,0009% error rate at 700Mhz indicates good overclock potential.
 
I tried 0 C cold air feed but fans don't go under 1920rpm and miner is running too cold, power usage went down 75W and APW3.. was louder than miner, 4750 Ghs.

[Finksy]

The 2000w PSU is not platinum, in fact it's silver but pretty close to gold. It's younger and stronger 2500w brother is gold rated, though. The 2880w beast is not actually platinum either, it is solidly in silver territory. I measured about 87% efficiency. Everyone confuses it with it's younger, slightly more buff 2980w brother the 39Y7414 is platinum, and much more expensive on ebay. We're all buying the 39Y7349 which has no official efficiency numbers anywhere that I can find as it's too old to have been put through the 80+ testing program. My own testing though, puts it at about 87% efficiency at 50% load.

Just curious, how did you measure efficiency?  And did you compare it against a known 80+ tested PSU to verify your methods?

[citronick]

Hi Guys - I need help on upgrading FW on the S7s to the latest ones.
The Bitmain support page lists a number of FW for different S7 batches and I need 2nd pair of eyes to double check which FW goes to which S7.

My current list as follows:

1. 4.73TH/v.1.32 (700Mhz) - Batch 8
2. 4.46TH/v.2.0 (600Mhz) - Batch ?
3. 4.66TH/v.1.7 (600Mhz) - Batch ?
4. 4.66TH/v.1.7 (600Mhz) - Batch ?
5. 4.73TH/v1.31 (700Mhz) - Batch 8

Appreciate any help.

Regards
TMT

For 2,3,4 use SD-S7-20151023-575M.tar.gz   (support 6 blade in one controller) or
this SD-S7-20151211-575.tar.gz  (3 blade only)

For 1 and 5 use SD-S7-20151211-700.tar.gz

Thanks Tupsu - I will get my machine up to date during my "New Year scheduled downtime".

The S7-700s are with latest FW and running sweet.
Keeping fingers crossed on the FW upgrade for 600s tomorrow midnight.

[wolfen]

Hi Guys - I need help on upgrading FW on the S7s to the latest ones.
The Bitmain support page lists a number of FW for different S7 batches and I need 2nd pair of eyes to double check which FW goes to which S7.

My current list as follows:

1. 4.73TH/v.1.32 (700Mhz) - Batch 8
2. 4.46TH/v.2.0 (600Mhz) - Batch ?
3. 4.66TH/v.1.7 (600Mhz) - Batch ?
4. 4.66TH/v.1.7 (600Mhz) - Batch ?
5. 4.73TH/v1.31 (700Mhz) - Batch 8

Appreciate any help.

Regards
TMT

For 2,3,4 use SD-S7-20151023-575M.tar.gz   (support 6 blade in one controller) or
this SD-S7-20151211-575.tar.gz  (3 blade only)

For 1 and 5 use SD-S7-20151211-700.tar.gz

Thanks Tupsu - I will get my machine up to date during my "New Year scheduled downtime".

The S7-700s are with latest FW and running sweet.
Keeping fingers crossed on the FW upgrade for 600s tomorrow midnight.

Been running dec 11 firmware since December 11 everything a ok.
When I switched I did notice more and smoother hashrate on the graph over 8 hours.
 130th to play with. You notice changes more at that level.

[philipma1957]

The 2000w PSU is not platinum, in fact it's silver but pretty close to gold. It's younger and stronger 2500w brother is gold rated, though. The 2880w beast is not actually platinum either, it is solidly in silver territory. I measured about 87% efficiency. Everyone confuses it with it's younger, slightly more buff 2980w brother the 39Y7414 is platinum, and much more expensive on ebay. We're all buying the 39Y7349 which has no official efficiency numbers anywhere that I can find as it's too old to have been put through the 80+ testing program. My own testing though, puts it at about 87% efficiency at 50% load.

Just curious, how did you measure efficiency?  And did you compare it against a known 80+ tested PSU to verify your methods?

I found it a little too hard to fully tease out the real numbers when I tested against my evga 1600 t2.

Here is what held me back the evga 1600 t2 sags in volts on settings over freq 600 so the highest freq I can test is 600 with the evga t2.   It provides just over 12.04 volts at the freq 600

When I test at freq 600 with the 2880w. The psu provide 12.18 volts   And this makes my test wrong.


This due to the nature of the s-7 string design it will spend the extra volts/watts with no gh gain.

So it is harder to really give it a fair comparison to the evga t2

Put another way if I set my s-7 to freq 600 with the evga 1600 t2 I can get  a 5 amp reading on my pdu at 240 volts.  If I set my s-7 to freq 600 with the 2880 watt I get a 5.9 amp reading on my pdu at 240 volts.

But the evga provide 12.05 volts and the 2880 watt provides 12.20 volts

The evga can run over freq 600 the 2880 can do freq 637
So do I want

4860 gh at 1200 watts maybe 1224 watts. Or do I want 5175 gh at 1416 watts maybe 1440 watts with the 2880 watt gear.

Mind you my s-7 is a batch 2 rated for freq 575

[Prelude]

The 2000w PSU is not platinum, in fact it's silver but pretty close to gold. It's younger and stronger 2500w brother is gold rated, though. The 2880w beast is not actually platinum either, it is solidly in silver territory. I measured about 87% efficiency. Everyone confuses it with it's younger, slightly more buff 2980w brother the 39Y7414 is platinum, and much more expensive on ebay. We're all buying the 39Y7349 which has no official efficiency numbers anywhere that I can find as it's too old to have been put through the 80+ testing program. My own testing though, puts it at about 87% efficiency at 50% load.

Just curious, how did you measure efficiency?  And did you compare it against a known 80+ tested PSU to verify your methods?

Simply compared AC input to DC output, that's all it takes. I've also tested 3 different server PSUs that have been 80+ tested. One of them was rated 94.5% @ 50% load (PFE1100), I read 94.1% consistently, for example. That difference can be due to my equipment VS their testing equipment, the PSU could be a little tired, or likely both. My testing mythology is sound, but I would of course appreciate someone backing up the numbers I have posted about the 2880W-er.

[OgNasty]

Yeah... Thread title needs to be updated. S7 is no longer 4.86TH/s or 0.25J/GH.

Should read:

ANTMINER S7 is available at bitmaintech.com with 4.73TH/s, 0.30J/GH

Since bitcointalk is not accessible from mainland China due to the GFW, could a mod update the title? I'm sure bitmain would in the pursuit of honesty, but they simply can't due to circumstances.   

The last two I bought arrived with 1 dead board each.  Maybe the title should read 3.1TH/s, 0.3J/GH?   

[Prelude]

Yeah... Thread title needs to be updated. S7 is no longer 4.86TH/s or 0.25J/GH.

Should read:

ANTMINER S7 is available at bitmaintech.com with 4.73TH/s, 0.30J/GH

Since bitcointalk is not accessible from mainland China due to the GFW, could a mod update the title? I'm sure bitmain would in the pursuit of honesty, but they simply can't due to circumstances.   

The last two I bought arrived with 1 dead board each.  Maybe the title should read 3.1TH/s, 0.3J/GH?   

Heh, sounds about right. My B8 had a board die within a week, so also running at ~3.1TH/s right now.

[Finksy]

The 2000w PSU is not platinum, in fact it's silver but pretty close to gold. It's younger and stronger 2500w brother is gold rated, though. The 2880w beast is not actually platinum either, it is solidly in silver territory. I measured about 87% efficiency. Everyone confuses it with it's younger, slightly more buff 2980w brother the 39Y7414 is platinum, and much more expensive on ebay. We're all buying the 39Y7349 which has no official efficiency numbers anywhere that I can find as it's too old to have been put through the 80+ testing program. My own testing though, puts it at about 87% efficiency at 50% load.

Just curious, how did you measure efficiency?  And did you compare it against a known 80+ tested PSU to verify your methods?

Simply compared AC input to DC output, that's all it takes. I've also tested 3 different server PSUs that have been 80+ tested. One of them was rated 94.5% @ 50% load (PFE1100), I read 94.1% consistently, for example. That difference can be due to my equipment VS their testing equipment, the PSU could be a little tired, or likely both. My testing mythology is sound, but I would of course appreciate someone backing up the numbers I have posted about the 2880W-er.

So you didn't take the Fan draw out of the equation I take it? That is part of the 80+ testing, as fan losses are subsidiary and not related to PSU efficiency, and the 3x60mm beastly fans that the 2880W PSU uses draw 1.68A or 20.5W each at full power (61.5W), which at the 50% load "testing" of 1440W would account for a margin of error of 4.27% alone.   You should check out the testing procedure for 80+ certification:  http://www.plugloadsolutions.com/docs/collatrl/print/Generalized_Internal_Power_Supply_Efficiency_Test_Protocol_R6.7.pdf

I would also love to see someone else do an accurate test on them, as I was informed that they were in fact 80+ platinum just as the larger 2980W PSU.  I have ordered 2x of the 2980W PSU's that I hope to put on a metered rack or ammeter and do some back to back testing vs the 2880W to see if there is an observable difference.  Still far from an accurate test, but worth looking into I think.

[VirosaGITS]

Yeah... Thread title needs to be updated. S7 is no longer 4.86TH/s or 0.25J/GH.

Should read:

ANTMINER S7 is available at bitmaintech.com with 4.73TH/s, 0.30J/GH

Since bitcointalk is not accessible from mainland China due to the GFW, could a mod update the title? I'm sure bitmain would in the pursuit of honesty, but they simply can't due to circumstances.  

The last two I bought arrived with 1 dead board each.  Maybe the title should read 3.1TH/s, 0.3J/GH?  

Heh, sounds about right. My B8 had a board die within a week, so also running at ~3.1TH/s right now.

That's pretty bad. Whats the delay and protocol here? You send back the black to Bitmain at your costs, then receive a blade within what delay? And then do you get compensated?

[adaseb]

Did anybody ROI on these miners yet? Especially if you bought the first batch and sold around the $500 highs recently?

[nhando]

Hi All,

Anyone have this issue where their S7 makes a really loud screeching sound from the FANs but if you place a small cardboard piece to funnel air differently then it would sound like other S7 where it's just rushed Air sound.  Please see image below.  I'm opening a support ticket as the sound is incredibly loud and annoying and over power the noise created by 3 other system by itself.

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

[Prelude]

The 2000w PSU is not platinum, in fact it's silver but pretty close to gold. It's younger and stronger 2500w brother is gold rated, though. The 2880w beast is not actually platinum either, it is solidly in silver territory. I measured about 87% efficiency. Everyone confuses it with it's younger, slightly more buff 2980w brother the 39Y7414 is platinum, and much more expensive on ebay. We're all buying the 39Y7349 which has no official efficiency numbers anywhere that I can find as it's too old to have been put through the 80+ testing program. My own testing though, puts it at about 87% efficiency at 50% load.

Just curious, how did you measure efficiency?  And did you compare it against a known 80+ tested PSU to verify your methods?

Simply compared AC input to DC output, that's all it takes. I've also tested 3 different server PSUs that have been 80+ tested. One of them was rated 94.5% @ 50% load (PFE1100), I read 94.1% consistently, for example. That difference can be due to my equipment VS their testing equipment, the PSU could be a little tired, or likely both. My testing mythology is sound, but I would of course appreciate someone backing up the numbers I have posted about the 2880W-er.

So you didn't take the Fan draw out of the equation I take it? That is part of the 80+ testing, as fan losses are subsidiary and not related to PSU efficiency, and the 3x60mm beastly fans that the 2880W PSU uses draw 1.68A or 20.5W each at full power (61.5W), which at the 50% load "testing" of 1440W would account for a margin of error of 4.27% alone.   You should check out the testing procedure for 80+ certification:  http://www.plugloadsolutions.com/docs/collatrl/print/Generalized_Internal_Power_Supply_Efficiency_Test_Protocol_R6.7.pdf

I would also love to see someone else do an accurate test on them, as I was informed that they were in fact 80+ platinum just as the larger 2980W PSU.  I have ordered 2x of the 2980W PSU's that I hope to put on a metered rack or ammeter and do some back to back testing vs the 2880W to see if there is an observable difference.  Still far from an accurate test, but worth looking into I think.

I'm replying in the other thread, this is off topic for the S7 thread.

[thedreamer]

Did anybody ROI on these miners yet? Especially if you bought the first batch and sold around the $500 highs recently?

I did, a week after I bought them. Made 30% profit on top from people that bought them. 

[E]

How is the noise level on the IBM 2990W PSU? Did they improve on the FAN used or same as the IBM 2880W? What else about it that makes them better than the IBM 2000 combo other than more power and connections?  I need to prepare for another batch of PSU order soon.

The fans are nearly identical. They have an updated model number but all of the specifications are the same. 2980w fans are four-wire, though.

The fans don't run balls-to-the-wall when the PSUs are in the BladeCenter H chassis and getting PMBus sweet nothings whispered to them... the 100% operation is a failure condition response, not design.

The 2880Ws have Delta PFB0612UHE: http://partner.delta-corp.com/Products/FANUploads/Specification/PFB0612UHE-F00.pdf and the 2980s are Delta PFC0612DE: http://partner.delta-corp.com/Products/FANUploads/Specification/PFC0612DE-F00(REV02).pdf