It doesn't work that way.  each algo is on a different port... so one setting per algo.... and the device would somehow need to know to try the next algo after disconnecting.  The password string just will kick it offline....  So this will not work.

That unit has a built in miner and web interface.

You would need to be able to log into it and modify the core scripting on it to attempt any sort of profit switching in it.  I assume you can enter pool/algo info in the web interface; if that's the case: no profit switching between algos.

Never heard of that miner.

Link to the manufacturer?

I was just keeping it simple.  feel free to move the decimal, but if the pool has the ability in the future to understand hashrate per connection in this way and take some sort of action;  it would be nice if it were a uniform known ratio.

OK....  I've put it off long enough....  Added link to this post in my batch file post as well.

How to "Normalize" your GPU/CPU for optimal profit switching
1.) create a batch file Similar to this:

2.) Run it.  Let it hash for 3-5 min.  Record the average hashrate that you are getting after this amount of time for the neoscrypt algo.

3.) Modify the batch file similar to this:

4.) Run it.  Let it hash for 3-5 min.  Record average hashrate for the argon2 algo as you did before.

5.) For the sake of being consistent the next algo (and so-on through all algorithms available for CPU/GPU) Which in this case would be yescrypt:

6.) Run it.  Let it hash for 3-5 min.  Record average hashrate as you did before.

7.) Now you have a list of algos and their hashrates.  The pool calculates the profitability number [generally] on a megahash (Mh) basis.  So our values need to be converted to MH from Kh or H/S.  So here's my values:
argon2=   12.1 Kh/s 0.0121Mh
yescrypt=   1.05 Kh/s 0.00105Mh
neoscrypt=   23.1 Kh/s 0.0231Mh

8.) In the password string for your round-robin of algos, you input the normalization value onto the identifier of what algos are mining like this:

NOTE: This is not necessary on a minercontrol config because you are using minercontrol to do the calculations for you and thus it doesn't rely on zPool's stratum disconnect to force the algo change to the miner unobtrusively and rather ingeniously.

In my batch you would simply modify my code above to suit your miner appname / algo name and run each one at a time... then edit this line in the batch to your numbers.   Note:  Blake algos are in Gh/s..  It should be easy to relate your results scaled over mine in the batch already.

**LMK if I made any typos**

Yesscrypt doesn't do #@$%# on a GPU.... well, nvidia anyways...   or hashrate is seriously under-profit compared to lets say X11.
I have found that yesscrypt is best on the new E-series cpu's with insane amounts of cores at higher freq...  But;  when I normalized on my I7 and my xeon X5650/X5365 machines;  Neoscrypt/argon2 were always preferred above all else.

Take the profitability in the corner with a grain of salt.  relate non-gpu algos with non-gpu algos only.  (Yesscrypt, m7m, argon2, neoscrypt)....   But the normalizations are magnatudes different than each other:

My Dual Xeon X5365 SLAED machine had these values [the values were scaled up a little in hashrate between it and the X5650 machine but no deviance same ratio of hash diff]:
argon2=   12.1 Kh/s 0.0121Mh
yescrypt=   1.05 Kh/s 0.00105Mh
neoscrypt=   23.1 Kh/s 0.0231Mh
So those MH values are what you multiply the current profitability estimate with that's displayed in the "Pool Status" pane on the site. (which [as I understand] is based on things like:  exchange price, pool luck, BTC exchange price, and hashrate/miners, etc)  That yescrype value is kind-of a false number... see why?  it's a thousandth, when the other's are 10-20x more when calculated to their respective lower number.   Yescrypt has been on the top since it first was added.   Only random exchange pricing causes one algo to take the cake, but that basically never happens.

so......



 for multi CPU profit switching it would look like this:

Now;  my batch will be modified at some point to do it for the cpu algos, but since there's only 3 I can use I leave it in the old "round robin" config. (haven't yet found a good working copy of m7m cpuminer for xeons and first gen i7's so its not here).  I can trim up the batch for this sometime if people think its necessary..

*edit* but yes;  You can run profit switching between cpu and GPU algos independently.

OK... looks like a lot has happened since I last checked in.

My day-job I am on-call for is a Audio/Lighting technician, and I just got back from a 4 day adventure in the woods for a private festival....   What an epic time that was.  We had a perfect high-altitude of the Pierced meteor shower.


Ok:  so lets see if I can get a solid answer here, please let me know if I missed someone else:


There's the link to the batch which is always updated in my signature.    In that post, are links to all of the miner apps that I am currently using to mine with my batch file.

I will say this straight up:  DO NOT MINE WITH A LAPTOP.  you will kill it.  Especially if it's an I5/i7 model as those little mobile processors run hot by themselves.  Trust me;  I have burnt out a Lenovo X201i, T420, and W520.  That's not a cheap price to pay either.  They flat-out do not have enough cooling to run 24/7 like miners do.  you would need it in a refrigerator with active cooling/convection.   Save the laptop and just build a PC.

For the PC side:  anything Core2-Quad generation and newer is highly recommended, though it will work on older/slower hardware.   The more cores your processor has, the better it will deal with trying to connect to many algos at once as my batch does.

For the software (remember I am only working with nvidia specific cards):  (referenced in my batch post)
I use ccminer 1.8.0 for all algos by default....  it's the "go-to" app.    I have found that it wont recognize the cuda cores and device identifiers on my GTX275 [test purposes]... and on a GTX480, it wouldn't work either.  For video cards that seem to not work with 1.8.0, I use the ccminer lbry or x11evo version which seems to have all the older core configs compiled in.  The newer one only seems to like 700+ series.  Anything 600 or earlier are hit and miss.


I'm guessing nicehash miner doesn't even see your video card... riiight? lol (assuming you tried that)
The 4600HD is not capable, I have yet to see one application for intel graphics mining myself.  You can search the webs, but don't hold your breath on finding one.

Id suggest buying anything in the 900 series if you are buying hardware to begin mining; and Id suggest a 960, 980, 980Ti, or any of the new 10 series.    The 750Ti cards are great, but only have so much efficiency with many algos.   They are very power conservative with others.  It's a waste of power/time to work with 500 series and earlier GTX models.... in my opinion, but for some people, hashrate is hashrate.  We have an encoding/render box in SF that has a GTX 650; and I still have it mine =)
\

For ATI:  I have yet to build a ATI powered miner (haven't bought an ATI card since they made the All In Wonder PRO in AGP format); so I can't comment on the nuances of setting up and configuring the software used for them.

I myself have two CPU's mining:
http://zpool.ca/?address=1PHSDYvVp6HpqtuUPocK41DrdeHbbezaeP
one machine has 2x 10 core xeon processors (X5650), and one machine with an i7-920 at stock clock.   If a LBRY block doesn't hit before you view my wallet, you will see how little those two machines earn in 24 hours...  

~0.00011/day for all processors mining....  Pretty minimal.... when you put together the notion that you are wasting more electricity on peripherals to run the processors than the processors take themselves... you will understand my meaning by it's almost not worth it.... but I won't talk you out of it cause i do it myself since my PC's is on mining with their video cards anyways.... plus I don't pay for power at home; included in my rental of property space.  You will need to google search "cryptominingblog XXX miner" where XXX is the algo(s) you wish to mine and download a copy of cpuminer.  Typically the newest algo's miner will have all or most of the older algos still compiled in, so you just have one version of cpuminer to run.  I have a simplified batch I can post up for that if/when needed.   But you have to disable virus protection.  Because of asshats with botnets;  antivirus software now thinks cpuminer is a bit of malware automatically.  Personally;  I solely point cpu's at neoscrypt.  It'a a power hungry algo, but the processor handles it best IMHO and it seems to plainly pay out the best.... FTC FTW.

Finally pool switching:  It's not something I have my batch configured for, and probably won't ever implement.  I see no need to.  I think its actually worse when you switch between pools, but this purely opinion.  Im here for the long haul.  The diversity and sheer profitability of zPool has my mind set on staying here.

OK.   I am done typing for now.   Hopefully this helps you figure out whats the hiccup in finding a starting point.   But the 3 basics are:
1) a nice up to date video card (With a GTX980 and a GTX960 SSC I got almost $10 in payouts on just Saturday just by mining LBRY)
2) a good working stable pc (I save old computers with multiple PCIE 16x slots for making miners.  I just save the motherboards and power supplies.  Some I even create a USB bootable Win7 OS to use as a "light" machine to sip even less power slower.  So when I get more cards from my business partner in SF, I just pull one out of my closet I have been saving and slap together a unit =)
3) a reliable internet connection

always glad to help you get on your feet if you want.... I post here fairly often, and as long as my job doesn't take me to nowhereland with no access to cellular/internet, I check in here multiple times daily.   Usually late night/Early AM PST.

I 100% agree.   Somehow I think its linked to that "garbage string" they seem to send to stratum (from what i've read)....

I wonder if someone compiles a fresh cgminer for them if it would fix the problem.

The strange thing is it seems to increase in severity with hashrate.

My S3's rarely saw downtime, and it was always happening on any pool anyways.
Ive heard of the S5's having issues staying connected on the pool
And worse with S7 Etc...

I wonder whats so different between your stratum and kano....

While this comotion was going on I decided to try and point an S7 back at zpool.   Started online for ~5 min before being kicked into retard mode where I can't even log into it.
Nicehash is working flawlessly.... so the issue is still pretty prevalent.

if people link me directly to built downloads for win7/8/10 x64, ill include them in my batch and update things;  but this is what I saw firsthand from current off-the-shelf options.  be sure to specify what algos are optimized per-link.

there will ALWAYS be improvements for one algo or another.

Just look up the cryptominingblog gtx 980 hashrate charts.     Totally not what they put out.


Grain of salt guys.

because exchange value on the coins changes......   it makes perfect sense.

:sorry for double post if it bugs anyone:

The readings were taken with a:
Device: Gigabyte GTX 750 Ti
Vendor ID: 1458:3667

It has no PCIE plugs, powered only by the 16x slot.  We use this little thin HP i7 machine to do video projections and video-mapping.
The is the one good thing about the 750Ti;  the TDP is phenomenal for the performance with multiple displays.

I was so tired when I finally got to re-coding and servicing the miners up there;  that I never changed the name of the miner; so I have a phantom miner masquerading as my one at home.   Oah well... lol.
My buddy is thinking of putting together a quad 1080 (no titan due to TDP) for doing the regular projection and video mapping to replace a machine with 4x 960 ssc cards..... 

But when those new pc's aren't at an event... they will be plugged in and hashing =)

No reference.  direct hashrates.  scrypt is useless on GPU so i left it out.

Ive been up for 42 hours straight now.....  barely got home 85Mi south.

And all hashrates taken with ccminer dev 1.8.0 as my batch specifies.


I will do a normalization on the 960 as well.   I wonder if the ,stats function of the pool can use the GPU type to choose a normalization automatically.....  from there we just need to provide the data to be put in tables for the pool to reference...   coding it in might be hard, but i'm not sure personally.

Long night in SF.  Got to the warehouse finally.

Here's some juicy tidbits:

EVGA GTX 750Ti normalization:
x17=1.8,
nist5=8.5,
x11evo=2.7,
sib=0.725,
x13=2.3,
x14=2.14,
x15=1.8,
x11=2.7,
qubit=4.2,
lyra2v2=6.2,
blakecoin=1.4,
blake2s=0.89,
neoscrypt=0.155,
decred=0.475,
skein=73.5,
lbry=44.2,
c11=2.74,
quark=5.0,


I'm tired.

Whomever is using my batch and has put some hash into my account a second time today:  I thank you:)

It may have been brief, but every little bit is appreciated!

I'm not sure what exchange zpool uses; crackfoo will probably comment.

I have seen the down ticks when the coins are sold at exchange as well.   Seems like a large percentage drop between block found and sold;  but... honestly with how crazy the markets have been, its no surprise.


I myself;  see people carry the distinction that BTC is separate in price to the dollar, etc...   but when anything bad happens to BTC to cause the price drop;  its not something bad happening to BTC 99% of the time, it's something happening focused on trust in others holding your coins or the exchange to fiat....   that's not a problem with bitcoin when its used as intended =)  I can tell how well my profitability is going via my 3 month grand-total [which is decreasing].   I think by now most if not all of my rentals should be off the list, and maybe even the larger machine up north I was mining into my personal account with as well.... So... Yeah.  Curious to see for sure.

Anyways;  since block find time (pool luck) has been below my normal batch... i'm curious at how well lbry keeps paying out in the next week to two weeks.

with unix the way to look at it back then was the shell was the compiler....

The problem is:

trying to something as simple in C as taking in data and using that input to be dynamic in the code... is haaaard in a batch.   This is why the C libraries PWN a lot of where batches fail =)

Ill just put it this way:  I remember dialing up my bbs, so I could fserv my first copy of Trumpet Winsock 1.0.....   Before then it was just telnet sessions to other BBS's when the functionality was added.     Growing up in silicon valley I always had access to computers and owned one at a young age.  I paud $125 for my first 2400 baud modem.  Took a LOT of work to save that $$$.

we should look into dropping some sort of legal hammer on them if its is at all possible.....

Im sure the community would band together over such nonsense taking place......