Quote from: glosed on February 04, 2018, 03:43:41 PM

Quote from: ylif on February 04, 2018, 04:37:09 AM

Quote from: glosed on February 04, 2018, 04:19:34 AM

Quote from: ylif on February 04, 2018, 03:54:27 AM

Quote from: glosed on February 03, 2018, 11:29:06 PM

Quote from: greyday on February 03, 2018, 05:17:34 AM

Quote from: glosed on February 03, 2018, 04:52:04 AM

Mining pool selection - I set up SiaMining at first and was getting around 80SC/hour. In the spirit of decentralizing and supporting good pools, I swapped over to Luxor, and my SC/hour was much smaller. After over 90 minutes it was estimating about 30SC, so I reverted back to SiaMining. However, I just refreshed the Luxor page and it is showing 362SC for just under 2 hour's work. Why does the number fluctuate so much, and why was that figure so delayed? Is that typical with Luxor, or is it due to the nature of PPLNS? Also, it seemed to take MUCH longer for my hashrate to reach ~800 on Luxor (30 mins or so) vs. SiaMining where I had full hash within a couple minutes. Or should I go with F2Pool instead? Haven't tried that one out yet.
HW error rate - People have been mentioning HW errors as "a lot of errors" or "an acceptable rate", but I haven't seen hard stats on what those rates actually are.
Is 0.01% too high? 0.1%? 1%? I'm currently at 0.006%...

I am jealous of your temps.

1) Mining pool--Luxor is having some sort of problem where their luck is currently at about 40%, so they switched over to PPLNS from PPS. Which means you get paid out in an average over time, but with their luck that low you're still going to make less right now (in theory). Siamining is PPS and currently paying out well. I'm with you though, I prefer Luxor and will switch back when whatever problem they're having gets sorted.

2) .01% is fine. .1% is ok. 1% is too much (opinions vary, but that's my standard).

3) Your cold air is your advantage, these things run hot. But I have mine overclocked (650 currently but usually 631). Overclocking voids your warranty, fair warning there. Your at-the-wall is your call; you can calculate profit with usage deducted at whattomine.com, it's usually pretty close. On a 20a 110 circuit, I'd suggest limiting wattage to 1700W for safety, you likely won't get anywhere near that though. HW isn't necessarily higher at higher frequencies, it depends. 631 drops more than 650 for me. It's more machine dependent.

Are HW errors at all related to the pool you are mining from? My hunch is no... but I was sitting in the other room and my miner shot up to 100% fan speed for no reason (when I checked the miner status PCB/Chip temps were like 30/40 so that wasn't the cause). But my error rate was nearly 3% which is way above what it has been to date (read: the 22 hours or so it's been running).

Also, it looks like there's something up with SiaMining, wondering if that could be the cause of the fan speed? Pool hashrate dropped like a rock. I figured it could be due to some miners going to AntPool, since https://siastats.info/mining_pools.html shows about 300TH/s that I don't think was there before. But then I noticed my own hashrate had gone down. To rule out that being the result of me getting a spike in HW errors or something else happening like it resetting and spiking the fanspeed, all/many of the addresses here: https://siamining.com/addresses are also getting cut in half...as if I needed a reason to switch back to Luxor, it does seem like their luck issues are getting better.

https://i.imgur.com/jJFgYJN.png

I found out that once you have overclocked, the hardware become less stable. Even you reture to the default 600, the hw error will suddenly jump in certain amount of time.(chips temp are <70) So I guess overclock is one way ticket, and there wont be anyway you can return to normal!

I was not overclocked. My HW rate typically fluctuates between 0.01% and 0.1%, but 3% is quite large. I feel like something may have happened to a lot of Antminers... my File System and Kernel versions are both from December, but it seems a bit odd that a big spike in my fanspeed and HW rate would occur at the precise moment there is a very noticeable dip in hash rates across SiaMining and F2Pool. Maybe it is coincidence... as greyday advised, I'm going to forget it, because it's Chinatown Tongue

Actually quite similar situation here other than overclocking. 0.1% is already too high according to the customer service. This may even because this A3 1st batch has not even finish its development with fully testing. Its all about win the time to conquer the sc1s market. Otherthan rebooting, i dont think there is anything we can do at this time as every minute now is so valuable in the future. And do not hesitate to control the hw, high hw will damage the hashboard no matter what your temp is。 i have already did some research on this hw error thing.

Thanks for that intel. Happened again to me this morning, right around 8am I was woken up by the fan speeds ramping to 100%. I just Save&Applied miner config to reset the machine, and when I did I was getting HW out the wazoo - over 100% on board 2, which to date has been the board that typically reports the least HW. Lots of x among the o's, and board 1 wasn't even hashing at 100GH/s. I tried another time and the same thing happened, which was strange because usually it's easy to fix. So I turned my unit off to go back to sleep for a bit and deal with it later. Woke up a couple hours later, fired it up, and she;'s cruising along just fine.

I'm glad that the issue is temporarily fixable, what I'm more concerned about is, what if this happens when I'm not at my apartment to make sure it's not running on big HW for hours on end?

I have come acrossed that start up issue too. What I realized is that the temp was too low （30-49）when i reboot it. I did some talk with people from bitmain，and they suggested me to raise the temp. So I turned off the 100%fan speed and it then started up just fine with higher temperature. （still less than 74）although the hw issue still happens, but it only happen like once in a few hours. The overall hashrate was not really affected by those sudden drops，as it returned normal quick. Now I just leave my a3 in this way，as anything else are fine. I will keep monitoring it until roi. I may try overclocking again in some days but Im just not sure.