Topic: ANTMINER S4 Discussion and Support Thread - page 133. (Read 301481 times)
So I worked with the bitmain guys just now on a workaround for the pool diff issue. Hopefully they'll post a fix for you to use. This should allow you to mine at other pools. While the hashrate will appear wrong when you first connect to them it should ramp up slowly to the correct value.
I'll be back in a bit now that's done to go through the last 15 pages or so to check for questions and to gather feedback.
Bitmain, are you planning on resolving these issues for the 2nd batch? I am immensely dissatisfied with my batch 1 miner.
S4 not satisfy me enough (plus btc prices)
Now waiting for L1
Now waiting for L1
kano.is after ~4h35m
ckpool.org:3333 ~4h40m
{"hashrate1m": "2.17T", "hashrate5m": "1.79T", "hashrate1hr": "1.6T", "hashrate1d": "582G", "hashrate7d": "44.1G", "workers": 1}
| Worker Name | Work Diff | Last Share | Shares | Diff | Invalid | Block % | Hash Rate |
| S4-3 | 1.37k | 5s | 4,682 | 6,268,852 | 0.730% | 0.084% | 1.64THs |
| S4-4 | 1.55k | 2s | 4,227 | 6,487,440 | 0.757% | 0.086% | 1.73THs |
ckpool.org:3333 ~4h40m
{"hashrate1m": "2.17T", "hashrate5m": "1.79T", "hashrate1hr": "1.6T", "hashrate1d": "582G", "hashrate7d": "44.1G", "workers": 1}
Great. Sounds good, but I suspect it's because our pools' starting diffs are 1024 rather than any intrinsic awesomeness of the pool itself.
Here try these as an experiment. I've set up extra solo pool nodes so the same conditions as the solo.ckpool.org (ie solo mine to your bitcoin address)
First runs diff of 2048:
s4.ckpool.org:3333
Second starts at diff 1 like ghash would:
s4.ckpool.org:3334
Bear in mind that ckpool adjusts diff VERY quickly so it may still work okay.
Here try these as an experiment. I've set up extra solo pool nodes so the same conditions as the solo.ckpool.org (ie solo mine to your bitcoin address)
First runs diff of 2048:
s4.ckpool.org:3333
Second starts at diff 1 like ghash would:
s4.ckpool.org:3334
Bear in mind that ckpool adjusts diff VERY quickly so it may still work okay.
2 on :3333 and 1 on 3334 … both seem to still work fine .
Thanks for testing, there is no need to keep those pool nodes online as they're identical to the solo pool just with a different starting diff so I'll be taking them down tomorrow but i'll leave them up in case others want to do similar tests. Note that you shouldn't mine long term on the diff1 node as it's not optimally set up for block finding. Anyone want to try one of these on a ckpool (kano's or my solo pool) and see how they fare? We don't have user selectable diff at the moment but it's just out of interest.
I just connected 3 , and they seem to work fine on this end …
Here try these as an experiment. I've set up extra solo pool nodes so the same conditions as the solo.ckpool.org (ie solo mine to your bitcoin address)
First runs diff of 2048:
s4.ckpool.org:3333
Second starts at diff 1 like ghash would:
s4.ckpool.org:3334
Bear in mind that ckpool adjusts diff VERY quickly so it may still work okay.
2 on :3333 and 1 on 3334 … both seem to still work fine .
This issue is that the S4 can't communicate it's need for higher share difficulty to the pool. Btcguild and ghash both work perfectly fine share difficulty manually set to 2048.
They shouldn't need to "communicate their need". It just means they're broken for lower diffs presumably because they can't keep up and just constantly overload when sending 15000 shares per second. Anyone want to try one of these on a ckpool (kano's or my solo pool) and see how they fare? We don't have user selectable diff at the moment but it's just out of interest.
I just connected 3 , and they seem to work fine on this end …
Here try these as an experiment. I've set up extra solo pool nodes so the same conditions as the solo.ckpool.org (ie solo mine to your bitcoin address)
First runs diff of 2048:
s4.ckpool.org:3333
Second starts at diff 1 like ghash would:
s4.ckpool.org:3334
Bear in mind that ckpool adjusts diff VERY quickly so it may still work okay.
Just one question in case I skipped it: Has the problem with the "underperformance" of the S4 been solved?
I am aware that the miners works with BTC Guild, but even after installing the posted "firmware" which updates cgminer I still only get ~110GH/s @ghash.io and other pools.
Thank you
I am aware that the miners works with BTC Guild, but even after installing the posted "firmware" which updates cgminer I still only get ~110GH/s @ghash.io and other pools.
Thank you
This issue is that the S4 can't communicate it's need for higher share difficulty to the pool. Btcguild and ghash both work perfectly fine share difficulty manually set to 2048.
Anyone want to try one of these on a ckpool (kano's or my solo pool) and see how they fare? We don't have user selectable diff at the moment but it's just out of interest.
I just connected 3 , and they seem to work fine on this end …
Just one question in case I skipped it: Has the problem with the "underperformance" of the S4 been solved?
I am aware that the miners works with BTC Guild, but even after installing the posted "firmware" which updates cgminer I still only get ~110GH/s @ghash.io and other pools.
Thank you
I am aware that the miners works with BTC Guild, but even after installing the posted "firmware" which updates cgminer I still only get ~110GH/s @ghash.io and other pools.
Thank you
We will try to connect 1 S4 your way. It should take at least a few more away before it can be done, but planning to point 1 S4 in about 4-6 hours.
PM you!
I guess I have my answer?
PM you!
Anyone want to try one of these on a ckpool (kano's or my solo pool) and see how they fare? We don't have user selectable diff at the moment but it's just out of interest.
My S4 is also stable at 2 TH/s as well.
What is your pool?It is solo.ckpool.org
I guess I have my answer?
@ yxt: Thanks for posting your p2pool results - it confirms what others have said here. The massive swings in hash rate as well as the spikes in DOA/Rejects is truly abysmal, at some points there are more DOA's than there is good hashes!
@ Bitmain: The S4 is terrible with p2pool. You should change your statement about them being compatible with p2pool until these issues, and others, are fixed.
@ Bitmain: The S4 is terrible with p2pool. You should change your statement about them being compatible with p2pool until these issues, and others, are fixed.
Thanks. Mindiff doesn't do anything yet on the pool. We're literally currently working on that.
Anyone want to try one of these on a ckpool (kano's or my solo pool) and see how they fare? We don't have user selectable diff at the moment but it's just out of interest.
My S4 is also stable at 2 TH/s as well.
What is your pool?It is solo.ckpool.org
I guess I have my answer?
I tested with stock firmware @ kano.is, 2 units ~1.75 - 1.8 TH/s (if I remember right), Minimum Diff was set to 2048 and 4096 under worker managment, but seems to have no impact.
I will switch back 2 units to kano, and one to ckpool now, update follows later)
EDIT: p2pool test:
Anyone want to try one of these on a ckpool (kano's or my solo pool) and see how they fare? We don't have user selectable diff at the moment but it's just out of interest.
My S4 is also stable at 2 TH/s as well.
What is your pool?It is solo.ckpool.org
I guess I have my answer?
Do not trust the Miner Status page, check your pools stats if possible
I have 10 S4 running.
All stable at 2 TH/s, no problem with the firmware our psu.
All stable at 2 TH/s, no problem with the firmware our psu.
My S4 is also stable at 2 TH/s as well.
Did you update cgminer from 3.12.0 to 4.6.1?
What is your pool?
It is solo.ckpool.org, but after updating the firmware, the miner status goes higher, but ckpool has gone lower.
In this picture it shows 3576.94 GH/S(5s), does that mean it went up to that for five seconds?
because since the firmware update the (5s) part seems to bounce up quite high? it didn't do that before the update.
http://i59.tinypic.com/1z4ctic.jpg
Jump to: