Scrypt profitability is now being adversely affected by the A4s

[fugju]

A4 136 mh went down to 1299 in price on ebay and the a2 stayed @ 1000 or better .is it because a4 are having issue and a2 are more stable.

1299$ + government fee is regardless too expansive. With 136mhs you got 4-5$ actually without electric costs and it goes everyday less. I think the real bang we gets, if Innocent do it like Alcheminer -> buy one get two.

[excelerator]

A bit of short term good news (3 day window) - LTC difficulty actually dropped and may be headed back to Nov 30 numbers. Probably overly optimistic, but hopefully we can begin to find the sweet spot for hashrate between the older scrypt rigs and the new A4s.
https://bitcoinwisdom.com/litecoin/difficulty

[excelerator]

The other way to drop LTC difficulty is to have a ~10% price drop like we're currently experiencing. That may be enough encouragement to have people switch off their old hardware.
https://coinmarketcap.com/currencies/litecoin/

[Steve Sokolowski]

The other way to drop LTC difficulty is to have a ~10% price drop like we're currently experiencing. That may be enough encouragement to have people switch off their old hardware.
https://coinmarketcap.com/currencies/litecoin/

Some of the difficulty increase is due to seasonal variation. When our basement lacked ventilation, we used scrypt miners to heat it. Even though the miners lost money, the electricity would have been used to power a dumb heater anyway, so it was worthwhile to run the miners regardless of profitability.

[excelerator]

The other way to drop LTC difficulty is to have a ~10% price drop like we're currently experiencing. That may be enough encouragement to have people switch off their old hardware.
https://coinmarketcap.com/currencies/litecoin/

Some of the difficulty increase is due to seasonal variation. When our basement lacked ventilation, we used scrypt miners to heat it. Even though the miners lost money, the electricity would have been used to power a dumb heater anyway, so it was worthwhile to run the miners regardless of profitability.

That certainly adds to the problem but we're in record difficulty territory with LTC and with the problems with A4s being resolved, this will likely skyrocket further reducing profitability. The "dumb heaters" will probably have little impact as they are low hash rate devices.

[excelerator]

When I look at https://bitcoinwisdom.com/litecoin/difficulty, it's a safe assumption that 500GH has been added to the LTC network since the end of September. Since there are no other new scrypt manufacturers, you can assume the bulk of that is the new A4. If we discount 20% of that is older/obsolete scrypt hardware being dragged out to heat dwellings in the northern hemisphere winter, then we are left with 400GH. That in turns equates to about 1400 new A4 miners. I've noticed that neither Nicehash or Prohashing have increased their share proportionally so this hash rate is going elsewhere.

If the top multipools like Prohashing, Nicehash and potentially Clevermining are not growing with the market then are the straight litecoin pools picking up all the business? If yes, then Innosilicon must be shipping 90% plus into the local market and dedicated data halls are growing.

[Steve Sokolowski]

When I look at https://bitcoinwisdom.com/litecoin/difficulty, it's a safe assumption that 500GH has been added to the LTC network since the end of September. Since there are no other new scrypt manufacturers, you can assume the bulk of that is the new A4. If we discount 20% of that is older/obsolete scrypt hardware being dragged out to heat dwellings in the northern hemisphere winter, then we are left with 400GH. That in turns equates to about 1400 new A4 miners. I've noticed that neither Nicehash or Prohashing have increased their share proportionally so this hash rate is going elsewhere.

If the top multipools like Prohashing, Nicehash and potentially Clevermining are not growing with the market then are the straight litecoin pools picking up all the business? If yes, then Innosilicon must be shipping 90% plus into the local market and dedicated data halls are growing.

Many of the first batch of defective A4s probably went to litecoin-only pools, because their owners mistakenly thought that they could earn more money there.

We haven't found a case where A4s are actually more profitable in LTC so far, because LTC has become so unprofitable. Switching coins does cause the A4s to lose money, but if they are auto-restarted, the hashrate decline seems to be be limited to 20%. We've been making 30% more than LTC consistently, so I think this is a messaging problem.

Clevermining hasn't been able to keep up with our profits for some time now, so an A4 owner who makes 20% over LTC at Clevermining may correctly deduce that there is no advantage of mining there over LTC with A4s. They may assume that all switching pools have that problem.

[excelerator]

So we need a third party review of miners & pools. Get someone who as cred on bitcointalk to do a matrix comparison and get it published on coindesk and/or cointelegraph. Only do A2s, A4s, and Titans. Only test against Nicehash, Prohashing and Clevermining. Everything benchmarked against LTC. Limit the measurement to daily average poolside hash rate and total return over 7 days. Would need all hardware in the same DC so latency is the same. It could be done and would be interesting.

[Steve Sokolowski]

So we need a third party review of miners & pools. Get someone who as cred on bitcointalk to do a matrix comparison and get it published on coindesk and/or cointelegraph. Only do A2s, A4s, and Titans. Only test against Nicehash, Prohashing and Clevermining. Everything benchmarked against LTC. Limit the measurement to daily average poolside hash rate and total return over 7 days. Would need all hardware in the same DC so latency is the same. It could be done and would be interesting.

Agreed - this is an excellent idea. We need to deal with the random issues that have occurred recently, so we can't do this now, but the holiday week will be much more open. We will need to identify who can perform this analysis.

[excelerator]

If we drop the requirement for a single DC, then we could in theory work with different individuals in different locations.

We would need 12 machines in total, each optimized for an individual pool (for LTC we would use litecoinpool). I can provide four Titans.

We just need two more people to provide the A4s and A2s.

Once we have this, then we just need to agree on the setup parameters for the testing (#xnsub, skipbcheck, etc.) and let it rip for 7 days.

We could configure them for anyone to watch via MRR for transparency once launched.