Heck. I don't even like Bitmain that much. I really don't care about what happens to the company. I only use the miners because they make the most money. As soon as bitfury gets their heads out of their ass's. I'll buy from them. If the A7 was any where close to being comparable. I would buy from them. I only care about making the most amount of money so I can continue to grow my mining business.
Yep that's pretty reasonable
My stats posts, which are supposed to have nothing to do with that
, are about the fact that a particular firmware, S9v1 has REALLY bad luck on the pool.
To the point where it is unlikely to just be "bad luck"
The S9v2 so far has been good (above expected luck), but like the A7, not a lot of data yet.
Yes but V2 firmware comes with it's own set of problems - at least in my experience and from what I've read with some others.
I've got 22 x s9's. I had zero issues with any up until first batch about 4 months into mining one board started producing xxxx's
Not a huge issue as scripts I've written would just reboot when this occurred and it would be good for random 15min - 15hours.
I had a couple others at same time start acting up in other units, but a downclock to 525 kept them solid.
I decided to try the autotune on a couple units. This one with the intermittent hashboard was one of two I tested.
It seemed to fix it. Or at least it showed as hashing away correctly as did the other unit. Other unit I updated was one that had to run 525.
Well autotune did help on this one and propped it up close to full hashrate without chips dropping. So thinking all was well I started to update more.
That's when i started getting a headache of issues. I updated 6 more units and seemed all good until about 24-48 hours into mining I had 4 units just stop hashing.
I could reach them via IP but they simply stopped hashing. I'd reboot them wait 40+ minutes and they'd eventually come back to the GUI and just sit there - not hashing.
I reflashed old firmware on these 4 - same result just didn't take as long before going back to GUI with no mining. Somehow the firmware was locking up in the PIC phase cycle and would just hang.
I then did remote reflash of firmware and still nothing. I ended up having to travel to the remote site and pull the 4 units. A hardware firmware reset on 2 of the controllers brought 2 of them back to life.
The other 2 would not recognize any blades until I removed everything from the chassis and re-seated everything and flashed them multiple times with the older firmware and I eventually got them to come back.
As soon as I left - 12 hours later one dropped again. This time was not able to resurrect it - and deduced the controller has gone bunk. So I've taken the three hash boards and piggy backed them into 3 working units for now until I can get a replacement control board. Since then I also learned that while the one unit with intermittent hashboard wasn't really hashing away at full speed. Pool side indicated that the board would drop, but gui hashrate never would fully reflect it. So I've since reflashed that back to old firmware so that my script could see the dropped board and powercycle as needed to get the most out of it.
As for other bitmain products - I had 4 boards go bad with 50+ s7's, and the early firmware releases were not good as major lockups would occur requiring power cycle and some dropped chips on the boards needing reboots. With the latest firmware though all the S7's I have almost never need a physical reboot nor drop chips. S4's I had never had anything but reboots needed - and fan replacements, they seemed to burn through the fans. S3 +'s nothing more than a random reboot from time to time.
Overall I think I would have been better off leaving most of the s9's on firmware v1.
Not sure if it makes a load of difference, but I have 3 or 4 different version of controller boards in these things.
Perhaps some boards do better with the v2 firmware and others don't. Not really certain.
