We don't have hard numbers. But based on the groundswell of negative reports on this forum, it certainly seems like the S9 has had more reliability problems than previous models.

I only bought a couple of S9s early on. One has a bad hashboard.

It's the first Bitmain hardware failure I've had since I started mining as a hobby. I had a couple of S3s, a Spondoolies SP20, and three S7s without a single failure.

I realize that my numbers are not statistically relevant. But from a personal and anecdotal perspective, the S9 has been less reliable.

That said, I still got to ROI with my S9s. Barely though. The board that crapped out happened about two weeks after that unit ROI'd.

Is bitcoinwisdom.com's difficulty page hung? Looks like it hasn't been updated in a few days.

Our youngest cat says:

eDonkey = 4.9

If ViaBTC is the only pool of consequence mining, and they have 127 PHs, then the probability (at the current difficulty) is that they'll find a block within 8 hours.

There's also rules in BCH that will reduce the difficulty mid-adjustment period. Maybe an expert (not me) can comment on what circumstances are needed to reduce the difficulty before the next adjustment.

Did you guys see the analysis of changes to the BCC difficulty rules that will allow large pools to manipulate the difficulty backwards, without waiting for the next adjustment period?

https://www.reddit.com/r/Bitcoin/comments/6q3joc/a_voice_of_reason_in_the_mining_community/

Assuming this is true, is this vulnerability an honest mistake of the devs who perhaps wanted mining to be effective during their first period as miners bail for the main BTC chain? Because if hash power follows BTC, the block rate for the first adjustment period of BCC would approach infinity, given the original rules.

Or was it intentionally designed to be exploitable by large pools like Antpool to maximize their acquisition of BCC block rewards during the initial pump and dump?

This change to the rules might make it interesting to have a thread to guess the difficulty of BCC. Not that I think it's a good idea to legitimize BCC, but wouldn't it be interesting to watch the difficulty go backwards during an adjustment period?

Regardless, the adjustment period for both chains after August 1st will likely be very interesting...

Trezor is going to provide a coin splitter. If you have a Trezor (like I do) then that means you can send BCC to an exchange without risking or touching your BTC.

There are a number of exchanges listed in this thread. Also here's an emerging list:

https://www.bestbitcoinexchange.io/segwit-bitcoincash-hardfork/

No. It supports a ton of algorithms, including Scrypt.

I agree that ProHashing does seem to be profitable, when it's not down.

I'm on zpool now. So far I'm seeing similar profitability but no significant downtime and no drama.

I have to agree. I wanted to like ProHasing, but finally gave up because they can't seem to keep the pool running reliably.

Some other nits that bother me about ProHashing:

• The pool web UI is overly fancy, prone to odd data failures (requiring logging out and back in again), and is mobile unfriendly.
• The idle miner email notification is broken and sends tons of false alarms.

I ended up moving to another pool.

Our girl cat says:

8.8 = eDonkey

If you put in bad URLs for your pools you can fire up a unit and the fans will run, albeit at a somewhat lower speed, without hashing.

I use a similar technique in reverse to cool my Antminers before shutting them down. Specifically, I change the pool URLs to something wrong (by adding an 'x' character to the domain, for example) then hit "Save & Apply". The result is the miner software restarts, but can't get work from any pools, so the fans run without the boards doing any real work. After a few minutes the output from the fans is significantly cooler.

I have to admit that I don't know if cooling them down before powering them off provides any actual value. But with the insane amount of BTUs these suckers pump out it feels like just shutting them off while hot is not a great idea.

Phil already answered, but just to clarify this is a pool specific setting. For Nicehash it's the "p" option in the password field.

For example, here's how I currently have Nicehash set up as my first pool (with a different receive address):


The above 0.3282 number works out to about 5% over the current SHA256 earnings. If Nicehash falls below that, it appears as dead to the miner and the miner switches to the next pool.

The pain in the ass part is that they don't allow me to specify a percentage. It's only a fixed number. So when there's a difficulty change, I recompute the profitability number and restart my miners with it.

I got tired of computing the profitability number by hand (using a mining calculator) so I now do it automatically using some Javascript in a Google Sheets spreadsheet. If anyone is interested, here's the GitHub page with the scripts:

https://github.com/edonkeycoin/gsbitcoinutils

I'm a little unhappy with the performance of blockr.io though (which I used for the scripts). I might switch APIs if I get pissed off enough.

I have my SHA256 miners on NiceHash with a profitability value. That last difficulty increase was so high that my miners switched away to a backup pool.

It's days like this that I wish NiceHash had listed to my request to optionally specify the profitability as a delta percentage rather than a fixed value.

Let's all hope the next jump goes easier on us...

Our 4 year old cat (her birthday was yesterday) says:

9.7 = eDonkey

I've seen 16% board level failure rate. I don't have nearly as many units as you, so this might not be statistically relevant.

My previous failure rate with a similar number of S3 and S7s was zero.

This new spike is scary. It's almost vertical.

Our girl cat thinks it's going way down this time.

-3.2 = eDonkey

Bump...

These almost make sense again with the resurgence in LTC value.

They're not on your list, but I have 13 Gridseed 5 chip pods. They're not sticks, but they're not blades either.

Let me know if you're interested.

Nice to hear from you suchmoon! Hope you're doing well.

I don't know if it's caching or not, but figured that would be an easy thing to try. Unfortunately the Linux tools installed by Bitmain are a bit sparse, so it wasn't easy after all.

A reboot has not helped either. The DNS servers are set by DHCP to point at OpenDNS. So I guess those servers have not been updated yet.

I'll try changing DNS servers. Or maybe more expedient would be to just use the IP address instead of the domain name for now.