I've tried that as well. I live in San Francisco and I have dedicated servers in a datacenter in San Francisco as well. I set up a full Bitcoin node on one of them and set up P2Pool. My latency was under 20ms and my DOA was around 2%. It was pretty much the most optimal setup that you can get without having them on the same network.
My node ( http://96.44.166.190:9332) is in LA. Looks like I have two folks there with a pair of S3s. Try pointing your workers there and see if you notice a difference. Also, despite all the talk going on here, I don't recommend messing with share size with + or /. Just use your payout address and see what happens. Once you know that's working, then try customizing it. M Switching them to your node right now with the default settings. Let's see what happens. I'll test them on your node until 7am tomorrow. My miner is the address ending with "osx". |
|
|
|
Why is this not documented? No wonder P2Pool isn't taking off. We need all critical code paths documented and logic explained.
On a side note, I refuse to believe that what I am experiencing is just variance and bad luck. It has been over 24 hours and I still haven't found a share with my 2 x S3s running at 1TH/s. There is definitely some sort of incompatibility between S3s and P2Pool.
While I love the idea behind P2Pool, sadly I won't be able to be a part of it until this is fixed. However, I am not hopeful that it will be fixed as let alone identifying the problem, it isn't even acknowledged yet. Pointing my miners to Eligius.
Try running a local p2pool node. Point your miners to it, and see what p2pool says for "expected time to share". M I've tried that as well. I live in San Francisco and I have dedicated servers in a datacenter in San Francisco as well. I set up a full Bitcoin node on one of them and set up P2Pool. My latency was under 20ms and my DOA was around 2%. It was pretty much the most optimal setup that you can get without having them on the same network. Expected time to share was about 5 hours and I haven't gotten any shares in over 24 hours with that setup either. Why is this not documented? No wonder P2Pool isn't taking off. We need all critical code paths documented and logic explained.
On a side note, I refuse to believe that what I am experiencing is just variance and bad luck. It has been over 24 hours and I still haven't found a share with my 2 x S3s running at 1TH/s. There is definitely some sort of incompatibility between S3s and P2Pool.
While I love the idea behind P2Pool, sadly I won't be able to be a part of it until this is fixed. However, I am not hopeful that it will be fixed as let alone identifying the problem, it isn't even acknowledged yet. Pointing my miners to Eligius.
I've got 5 S3s pointed to p2pool and have found 11 shares in the past 24 hours. There's nothing wrong with how they function with p2pool. I wish you the best of luck on Eligius. Wizkid's got a good pool there. While Eligius is really good and I am fine with switching to them, I really want to stick with P2Pool if possible. It is the "right" way to mine in my mind. I have two S3s and you would expect them to have found around 4 shares since your 5 found 11. Variance, bad luck etc. but still 0 shares? That doesn't make much sense to me. We can't just decide not to even try to debug a distributed system because it has a random component. It's like saying it's not a bug, it's a feature. I agree it sucks that you haven't found any shares but what exactly would somebody debug? If every S3 suffered the problem (like the S2's documented loss of hash rate) then there would be something to look into. If your miners are reporting good hash rates with low errors, low rejects and low DOA then they're working properly. Didn't you previously post that the same hardware was finding shares a few weeks ago? If that wasn't you then I apologize. My point though is that the most likely answer is simply bad luck. There just isn't enough evidence to point to problems with the hardware. If P2Pool was better documented, I would be able to form some sort of hypothesis but I can't due to the lack of documentation. I know that P2Pool server essentially "lies" the the miner with psuedoshares so that it keeps mining. That's why I don't think the miner-side statistics would be very helpful in this case. I could look into the code and try to figure out what might be going wrong but that's what I do all day with my job anyway.  And you're correct, that was me. When I first started mining with my S3s on P2Pool, I found four, maybe five shares in a row but I pretty much never found any other shares after that. I am running on the latest firmware by the way. I am not sure if others upgraded to it. It could be a problem with that firmware. As far as I can tell Antminer S3s are running a modified version of "cgminer 3.12.0". This seems to be outdated and I might try cross-compiling the newest version. However, that lacks the Bitmain modifications and it will be less than ideal. |
|
|
|
Why is this not documented? No wonder P2Pool isn't taking off. We need all critical code paths documented and logic explained.
On a side note, I refuse to believe that what I am experiencing is just variance and bad luck. It has been over 24 hours and I still haven't found a share with my 2 x S3s running at 1TH/s. There is definitely some sort of incompatibility between S3s and P2Pool.
While I love the idea behind P2Pool, sadly I won't be able to be a part of it until this is fixed. However, I am not hopeful that it will be fixed as let alone identifying the problem, it isn't even acknowledged yet. Pointing my miners to Eligius.
Try running a local p2pool node. Point your miners to it, and see what p2pool says for "expected time to share". M I've tried that as well. I live in San Francisco and I have dedicated servers in a datacenter in San Francisco as well. I set up a full Bitcoin node on one of them and set up P2Pool. My latency was under 20ms and my DOA was around 2%. It was pretty much the most optimal setup that you can get without having them on the same network. Expected time to share was about 5 hours and I haven't gotten any shares in over 24 hours with that setup either. Why is this not documented? No wonder P2Pool isn't taking off. We need all critical code paths documented and logic explained.
On a side note, I refuse to believe that what I am experiencing is just variance and bad luck. It has been over 24 hours and I still haven't found a share with my 2 x S3s running at 1TH/s. There is definitely some sort of incompatibility between S3s and P2Pool.
While I love the idea behind P2Pool, sadly I won't be able to be a part of it until this is fixed. However, I am not hopeful that it will be fixed as let alone identifying the problem, it isn't even acknowledged yet. Pointing my miners to Eligius.
I've got 5 S3s pointed to p2pool and have found 11 shares in the past 24 hours. There's nothing wrong with how they function with p2pool. I wish you the best of luck on Eligius. Wizkid's got a good pool there. While Eligius is really good and I am fine with switching to them, I really want to stick with P2Pool if possible. It is the "right" way to mine in my mind. I have two S3s and you would expect them to have found around 4 shares since your 5 found 11. Variance, bad luck etc. but still 0 shares? That doesn't make much sense to me. We can't just decide not to even try to debug a distributed system because it has a random component. It's like saying it's not a bug, it's a feature. |
|
|
|
|
Why is this not documented? No wonder P2Pool isn't taking off. We need all critical code paths documented and logic explained.
On a side note, I refuse to believe that what I am experiencing is just variance and bad luck. It has been over 24 hours and I still haven't found a share with my 2 x S3s running at 1TH/s. There is definitely some sort of incompatibility between S3s and P2Pool.
While I love the idea behind P2Pool, sadly I won't be able to be a part of it until this is fixed. However, I am not hopeful that it will be fixed as let alone identifying the problem, it isn't even acknowledged yet. Pointing my miners to Eligius.
|
|
|
|
|
While this can be attributed to variance, I find it hard to believe that it is just variance.
When mining at Eligius with my S3s, I can easily get shares as good as 260 million within a day while I can only get shares around 4 million if I am on P2Pool. Something is wrong here.
|
|
|
|
I have been mining on P2Pool lately and even though I had great payouts initially, I haven't been finding any shares lately.
I have two Antminer S3s and I have been thinking if it is a bad idea to have both of them mine to the same address. Wouldn't that make the P2Pool set a higher difficulty for both of them? Should I be mining to different addresses with them?
The pseudoshare will be higher, yes. The shares that count will be the same regardless. M So the actual share difficulty is actually not based on the hashrate of the miner. Only the psuedoshare difficulty depends on the hashrate. Is that correct? Are psuedoshares valued at all? Because even though my node says approximately 6 hours to a share, I don't get any shares even when I run my miner for over a day. Pseudoshares only make your hashrate look better on the graph. The actual share chain difficulty is based upon the entire hashrate of the pool. Keep in mind luck (aka variance) plays a role here. When it says approx 6 hours, that's on average. You'll have some that are less, some that are more, but your luck has to be pretty bad to not get one within 24 hours. I've got two folks on my node that look like they're using two S3s a piece. They're chugging right along with shares, so S3s seem to be working well. M I don't think my miners can be this unlucky, though. No shares in this long might indicate a problem. I remember a couple of weeks ago, when I was mining on the same node with the same hardware, I was getting shares constantly. Sure, variance but this seems too much. I am doing stratum+TCP, should I try http? My DOA is about 3% so no problems there. My latency is no more than 80ms as well. |
|
|
|
I have been mining on P2Pool lately and even though I had great payouts initially, I haven't been finding any shares lately.
I have two Antminer S3s and I have been thinking if it is a bad idea to have both of them mine to the same address. Wouldn't that make the P2Pool set a higher difficulty for both of them? Should I be mining to different addresses with them?
The pseudoshare will be higher, yes. The shares that count will be the same regardless. M So the actual share difficulty is actually not based on the hashrate of the miner. Only the psuedoshare difficulty depends on the hashrate. Is that correct? Are psuedoshares valued at all? Because even though my node says approximately 6 hours to a share, I don't get any shares even when I run my miner for over a day. |
|
|
|
|
I have been mining on P2Pool lately and even though I had great payouts initially, I haven't been finding any shares lately.
I have two Antminer S3s and I have been thinking if it is a bad idea to have both of them mine to the same address. Wouldn't that make the P2Pool set a higher difficulty for both of them? Should I be mining to different addresses with them?
|
|
|
|
|
Payout queue seems empty and even though I should be paid, I am not getting paid right now. All the block rewards found right now are going to a single address. What's going on today?
|
|
|
|
|
Is anyone mining on P2Pool with their Antminer? I was getting great results but now I can barely make earn any shares. Is the firmware still incompatible?
|
|
|
|
Yes "warranty' have not much value in the mining hardware.
41 - 42 was my temperature peaks at stock clocks.
I don't have any link to the thermal grease for now.
I'm searching for it to buy the same brand or the alternative one having >3 conductivity.
But most of the products I'm seeing is of conductivity 1.
I will post here the link if I find it or some else who has experience in doing this will support here.
Interesting my miners are all running in the same temperature range like yours and I have similar performance to yours maybe this will help me too. Still though our temps are half of the max 80C but I guess its averaging the temp so maybe some things are hotter than others.... My miners run at those temperatures as well, if not more, and I haven't been having low hashrate issues. In fact they both are performing at low 500s. I don't think the temperature is a symptom at all. |
|
|
|
All 4 of my Batch 3 units came stock 218.75 and happily sit at 442 after 12 hours, Is it worth bumping it upto 237 or 250 or not worth the hassle?
If you are lucky you might have gotten machines that can hash at ~500GH/s on 250, or ~480GH/s on 237. Worth a try. |
|
|
|
Why is everyone still beating this dead horse.? It's been pretty well established that "HEAT" is NOT the cause of the S3 problems, it poor design and possible sub-par components.
We've already read that some users have put their S3's in front of A/C units and experienced no improvement.
Now we have pages of people trying to put heatsinks on chips that dont need them.
Lets face it , Bitmains first performance figures were "optimistic" and probably thrown out there in a hurry to get the machine to market, and after further testing they realized that they made a mistake.
That's why they gave us money back, not enough but a good gesture, not many companies these days would do that.
If we are into Batch 4 and possibly batch 5 with no hardware changes then I think you can safely assume that it is what it is and probably cannot be fixed.
If there was a fix to be made, I think Bitmain would have done it .
This is not a problem that firmware up grades will fix either.
JMHO
Bob
I tend to agree with you. I believe the stock setting should be 212.5 with 430gh and 328 watts on a gold psu. They now sell them at 218 with 441gh and 340 watts on a gold psu. >>>>>>>>> I feel this is not stock but a small overclock. The real mystery is the s-3's that do 480gh or 500gh and freq 237 or freq 250. It is definitely a mystery. My first two from B1 can stably hash at: 480GH/s at freq 237 510GH/s at freq 250 I'm using EVGA Supernova 750w and Corsair CX600 to power them. They reach to temperatures at around 45C. |
|
|
|
Why are you folks in the USA picking UPS over DHL? DHL is way better at handling customs. I ordered a bunch of Zeusminer stuff (half of it was free because of their crazy coupons but you still had to pay the shipping) and it always came through DHL like in less than 48 hours from China to my crib. DHL also lets you pre-sign if you are not around.
I would always pick DHL if possible when shipping from China if you are in the USA.
We weren't given an option, Bitmain just shipped them with UPS. |
|
|
|
7.7% refund posted to my account just now. Pending....
Transaction now complete!! Mine still shows "Init". Scratch that. I just checked my wallet, it has been sent. It is "unconfirmed" right now. Refunds are sent from 13WHCm8DztmNGZ7XX1tbhUDss4WgRMVnrj. |
|
|
|
Can anyone suggest if this is a good PSU for 1 S3?
Corsair CX750 Builder Series ATX 80 PLUS Bronze Certified Power Supply.
That one should do it. I don't know how many PCI-e outputs it has but two should be enough if you aren't overclocking. I am currently running two S3's with the following configuration: EVGA 750W 4 PCI-ex goes to one S3 EVGA 750W 2 PCI-ex goes to the slave blade on the second S3 while Corsair CX600 Builder Series supports the master blade of the second S3. Both of them are reliably running at 500 GH/s. |
|
|
|
<VENT>
I was batch 1
ordered 18 S3 miners
One S3 had 14 chips marked x and eventually -
I contacted bitmain via the new messaging system on their website and patiently awaited a reply for the next 7 days.
In the mean time another unit stopped hashing although I could view the web gui. Just no chains.. gone.... Fans at full tilt. Tried re-seating cables.. nope.
So, I took the control board out of the unit with the 14 dead chips and put it in the unit with no chains... Success, a functioning S3 again.
Contacting Bitmain on these forums I explained the situation as was directed to email them. Which I did, explained the whole ordeal and asked for advise on what to do with malfunctioning unit.
Flash new firmware was the reply, dont check keep settings. Failure. I informed them so. Silence...
Over my time I have made 24 seperate orders from BITMAIN totalling ~50BTC.
First time ive contacted them to say, "Hey, this ones broken".
Im not big on public outcries, however, on this forum-thread I note another member who had controller board problems and it wasnt until he made it public that action was received in the form of an RMA.
So, here I am
</VENT>
BITMAIN really has to work on their customer relations. Missing refunds, careless customer service... |
|
|
|
Over the past days I've had a batch of Antminer S3's mysteriously power off. Power supplies check out fine, they just shut down, no overheating, or circuit issues, just off. In order to wake them up I have to recycle the power supply. It's not isolated to one Antminer or one power supply, but multiple units in different areas, different power, etc. Different pci-e cables, different types of power supplies, doesn't matter.
Ideas?
You might be overloading the PSU. Overclocking would do it, or if your PSU was barely supporting the S3s through its PCI-ex, you might have that issue. When that happens the PSU needs to stay disconnected for a bit and then it will start working when connected back to the wall. It is because of a failsafe within the PSU. |
|
|
|
Has anyone received their refund yet? I mean we had an hour to pay, They've had weeks.. An ETA would be nice @Bitmain
Refunds  what on earth for. Unless I missed something was Bitmain refunding for prices or something on offer for units ? or people just asking for refund of unit? Yes, you definitely missed something. Since the miners were performing below the initially advertised hashrates, buyers were given the option for a 7.7% refund or 10% coupons. Coupons are issued but the refunds aren't yet. |
|
|
|
|
What's a good ratio for HW/Accepted and Discarded/Accepted?
|
|
|
|
|
Show Posts
