unlikely it will drop like that. The cloudmining contracts might end, but the hardware will still exist. It may likely just be used for CEX's private mining.

I expect a 0% diff change because a lot of 1w+/gh gear will shut off but be replaced by newer S5/SP20 gear

might see a negative tbh. if anyone was considering shutting off hardware durin ghte last few days, a 7% drop in mining income might be the final blow.

yeah that bummed me too - the heat of 1Kwh (~$0.15) is equal to about $0.02-0.025 worth of natural gas, so savings are marginal

I have virtually zero experience compiling stuff on the RPi via shell. a bit of example code would really help. AFAIK minepeon runs cgminer in a specific folder, would this work:

**Fresh install minepeon
*This runs 1.7 to 2 stable, I think we should compare charts to bfg to see if it matters

sudo pacman -Syu
sudo pacman -S git base-devel ncurses curl libusbx jre7-openjdk-headless
**Hit Enter
*Hit Y
cd /opt/
sudo git clone https://github.com/FireWalkerX/cgminer-bmsc.git
cd cgminer-bmsc/
sudo ./autogen.sh --enable-bmsc --enable-bitmain
sudo rm /opt/minepeon/bin/cgminer/
sudo make install
cd ..
sudo mv cgminer-bmsc /opt/minepeon/bin/cgminer
**edit mine peon startup settings
/usr/bin/screen -dmS miner /opt/minepeon/bin/cgminer/cgminer --bmsc-options 115200:20 --bmsc-freq 0881 -c /opt/minepeon/etc/miner.conf
**Reboot

tried with one board only, and put pins 3,4,5 as TX,GND,RX to a CP2102. minepeon v0.2.4.6

tried restarting a few times, rebooting, and setting it to use cgminer rather than the default bfgminer. no luck yet - whats wrong? the 3 UART wires seems to be in snug contact with the S1 pins

plugging in the CP2102 results in a brief flash of the hashing board's LED - nothing beyond that

Hopefully the first post of the thread can be updated one day - 36 pages is a lot to wade through. Ive got my undervolted hardware running pretty well using the stock control boards, and now have an RPi w/ 8GB class10 card ready to go. Will the new minepeon be capable of running my boards? Also, how much power is drawn over USB? will a 10port hub work unpowered if i connect 10 boards to it, and if not will a 1A hub be sufficient?

ps: I mucked up a unit with hex value 0e02 - should have been 187MHz roughly, but i think it killed either the boards or their controller. hoping the RPi +cp2102 will solve what went wrong

diff is creeping down nicely, might be in the 7.5% range

wouldnt surprise me if GAW has that much hashrate. look at the paycoin scam - wouldnt be surprised if they were a major source of hashrate rentals on nicehash to make money on the paycoin nonsense. (sure they could mine XPY themselves, but they already have 12mil of them they are flooding the market with - renting out thier cloudmining to mine thier own worthless coin for BTC seems like a clear win)

Interesting. How did you decide to use those frequency numbers? I would imagine that at a lower voltage it may be necessary to reduce the frequency at a lower rate.

the packing is different now - the first bunch of S5  units shipped in foam/bubble wrap just like the S3. This packaging is likely lighter and more environmental

Bitmain has a large facility that could handle several PH - and its pretty likely that they test the S5 units (even briefly) before shipping them. similarly, SP-T has been shipping thousands of the SP20 unit, and likely performs some testing before its boxed and shipped.

there is usually some big varience swings in the difficulty calculations, but i think theres also smaller ripples caused by manufacturers bringing several hundred units online for testing, running them overnight, then shutting them down to ship a day or two later.

If you have a cold air source, or can handle a lot of noise, 412.5MHz seems to be about the limit before hashrate starts nose-diving. Board temperature around 65C doesn't seem to be the problem. Peaks out around 1.32-1.35TH, and at 425MHz is down to 1.25TH

6TH seems abnormal for an SP35 - that is roughly a 3800-4000W DC12V draw, and the SP35 contains a pair of PSUs that are rated to give only 1600W DC each, for about 5.5TH/3200W (~3550W at wall). would love to know more, because if it can actually drive itself to 6TH i want some to go alonside my SP31 units (limited to 4.68TH at 208V)

+1. The S5 is perfectly capable of running with <0.05% errors as high as 60-65C, perhaps higher (i wouldnt suggest pushing any higher though) The PWM settings are the culprit here and Bitmain needs to follow the example of how SP-T controls its fans.

At 7V, in a 21C room, and with a few clear packing tape strips over the top of the device to better tunnel airflow - I can run 350MHz/58C or 362.5MHz/61C

the varience is pretty big, dropping back down to expected levels right now, hopefully that will hold for a few days and the jump will be around 7-8%. Im guessing a few things are happening this timeframe simultaneously:
1) Paycoin miners returned 50PH to bitcoin. however the price drop and difficulty spike likely made a lot of these machines realize they are not profitable for BTC anymore
2) There are probably a lot of people mining at >1w/gh unprofitably who were doing so in hopes of long-term price. With those hopes shattered <$300 the hardware is turning off.
3) low BTC price means little buying power for new hardware.
4) the spike could be the 24-72hr testing of bitmain or SP-T hardware before it ships to customers. not likely, but it could account for 5-10PH swings as stuff is plugged in, setup, tested, turned off, then boxed to ship and the cycle repeats for another batch.

5) selfish mining / block withholding. I'm not sure how big an issue this still is right now, but a few mid/large-sized farms that mine at a pool could withhold solved blocks from the pool in order to be paid ~99% of the normal payout but effectively deprive the network of several TH or even PH of solved blocks. doing this holds down the difficulty, and unless the pool is smart enough to check if there are (x blocks)/(y shares) it could go unnoticed and a pool would appear to just be having bad luck

I think we are still going sideways in difficulty until price goes back over $400, or at least $350. likely 8% this time, then 2%, and then anything in a (-1)-(+4)%/jump from there as any new hashrate requires the removal of older gear.

^possible - but it depends on a lot of things:
1) lots of investors were getting involved when btc was >$500
2) does Ken have any assets left in the business, or have all assets been sold and there is only an unpiad debt?
3) is there reason to beleive that ken's personal assets (car, home, bank acct) were funded from ACTM/VMC, or should otherwise be liquidated to compensate the investors.
4) Is ken being seen as an owner of a bankrupt business, or a fraudster running unlicensed securities that stole from unsavvy investors

seems that the pencil mods increase in effect slightly over time, Ive had to adjust all my frequencies about 6-12MHz lower to minimize 'x' chips.

to that extent, I present the hex value for asic-freq frequency 187.5MHz .......    0801  testing it right now to see how it works out

probably not soon - using the S5 blades could be too much heat for the heatsinks of a C1 (1200W+ vs the C1 was 800W)

IMO, would rather see an option that is more like the S2, many smaller blades running efficiently in a server case.

I sold the SP20 units, and dont have a killawatt. My estimates are based on wire/PSU temp (the CS650M gets quite warm over 580W, and very hot in spots at 600W+, same for the PCIe cables which get barely warm at a ~300W draw each), and the DC12V draw readings of the SP20.

for the SP20 my settings were similar to most people getting 1.15-1.3TH from the device at low fan speeds. As i recall;
start: 0.635 max: 0.655 wattage: 155 fan: 10% (cold air at inlet)
running like this said 150W draw per PCIe (so presumably 600W 12VDC)

the 661W number must be at the wall - the S5 spec indicates a draw of about 560-600W on 12V DC. I pushed a unit to 287.5MHz before it tripped my CS650M (gold ATX with 612.5W 12V rail - trips somewhere around 630W) and ran at 362.5MHz 24/7 for several days without issue, indicating about a 600-610W load on the 12V rail

a 1200W (GOLD - most are when over 1200W) PSU should be able to run two S5 units at stock, or if necessary slightly below. A bit of overhead is better (<95% load is a good target for the PSU), so a 1300W PSU would give you that overhead and likely allow a bit of overclocking to achieve 1.21TH/S5

dont o for a PSU any less than gold. The difference between gold and bronze at 90% load is about a 3-5% difference in efficiency (or ~50W on a 1200W draw) If this is ~$3/month more for electricity the savings will cover the higher PSU cost within a year or two, and the supply will be better quality and resale value