Topic: Hash Auger 2.9.7.5 Mining Manager and Switcher for NVIDIA GPUs - page 24. (Read 8755 times)

I will need some more information to help me troubleshoot the issue with copying the OC settings from one device to another.  In a quick test, I set OC settings for a couple algorithms on GPU0 and then saved the device settings. I then went to GPU1 and selected GPU0 from the list of devices to copy benchmarks from. A message box asked me for confirmation. I pressed yes and all the OC information copied over for each algorithm I had changed. The only thing that didn't copy were the intensity values, which I am going to fix for 1.8. I even did a few variations, such as reversing the cards, not saving the device benchmarks before copying and not enabling OC settings on the device that the benchmarks were being copied to. All the OC settings copied over in each of my tests.

I could use a little more information from you to help identify the issue. For example, did a message box appear after you selected GPU0 from the list? Did the benchmark hash rates change after you copied the settings over or did the entire list of benchmarks appear unchanged?

Also, I wanted to add that the software will use the OC settings on each device's Hardware tab if there isn't custom Overclock settings for a particular algorithm.  This allows you to set a default overclock profile that will apply to most algorithms without having to type in the same numbers for each algorithm. Then you can just enter custom settings for algorithms that should use different overclock settings. For example, you could set a default OC profile of +100 GPU and +300 Mem at 90% power limit on the Hardware tab and settings of 0 GPU and +100 Mem at 85% power limit for Raven.  The default settings would then apply to every algorithm except for Raven.

in the setup section, on devices, when I customize the "ALGOS" with your OC, then I save it. But copying that configuration to the other cards does not work.

I save the OC configuration for each protocol, for example in GPU0, and then in the next I tell it to copy the GPU0 data and it just does not do anything.

It is a lot of work to do one by one the 6 cards of a rig, and configure the OC protocol to protocol. So the copy function does not work.

Another suggestion, there could be a more advanced benchmark, where it tests several intensities automatically to choose the one that offers the best hash performance. It would take longer, but it could be optional.

It is that, if all bots nemosminer, nicehash, hashauger etc ... always change to the most profitable, after a few minutes it stops being profitable and changes. In other words, in addition to making the profit fall, you have many more changes and more time lost between changes of coins and miners.

Maybe entering the second most profitable, fewer miners will enter, more time I will be extracting and maybe more benefit.

Is that how is this soft and others, in the end we ALL agree on the same coins in a short time, and we do not make it profitable because there are too many miners, with which the profit plummets in a few minutes and the Change to that currency in the end has not been profitable.

I personally use a 12% change in benefits, to have fewer changes.
When everyone is going to undermine the most profitable, I personally prefer the second currency, because it will be more real profit when fewer miners enter in a short time, at the same time will make the first currency hold more, and that way everyone earns more , those that go to the first currency and those that go to the second currency, because it diversifies the HASH between two or three coins.

You also have to control it, because it may be that the second currency is 50% less profitable than the first, but surely some option to control that could be added.

That is a good idea to give us the option mine the 2nd and 3rd most profitable(more so for gpu as I feel like cpu miners is a smaller group).

I went out last night and when I came back I was getting pool errors with AHashPool(gpu, cpu was off as it was below the minimum price), saying something to the effect software can't connect to pool. I stop then pressed mine again same thing(did it a few times to check). I restart the app and chooses zergpool. Maybe it was an app thing, or maybe it was something with the pool, but I thought I'd mention it. Thank you.

Thank you for trying the software and your detailed feedback.  All of your ideas are interesting and some of them are features that I am already considering in the near future. They most likely won't all be included in the next major update (1., but I am hoping that you will find the next few major releases (1.8, 1.9, 2.0 etc) to have a significant increase in versatility and flexibility.

First of all thank you for your program and its simplicity, I love it.
I am going to suggest a couple of functions to increase your versatility.

1.- I can have a list of currencies configured with your pool and wallet and that the program only makes the change between them. Right now you can not, now you can switch between algos, or mine a currency. But I would like to have 10 coins defined with their pools and wallets and just watch and switch between them. That would be a great point.

2.- Exchange option, that instead of entering the most profitable of the moment, enter the second or third. Because when a currency is profitable, suddenly thousands of bots like this enter that same currency and pool and kill their profitability, but if I enter the second or third that would not happen.

And a suggestion, I know it will be laborious, but implement Suprnova, the problem is that every thing has its url, login and pass. But it is one of the best groups there are for serious currencies.

Thanks for your great program and I hope you keep in mind even if it is only the first suggestion that I find VERY INTERESTING

If they use XmRig, then the miner support is there, but I still need to find a pool that mines Graft. I'm working on including support for some pools that mine lesser-known coins, so maybe one of them will support that coin.

Blazepool really put in a lot of effort to modernize their UI and as you said, they appear to be very responsive to user concerns. Definitely a pool to watch.

The Price Spike Limit is based on the price calculated by your device benchmarks and the pool's price estimates.  The issue with Tribus is that the actual hash rate when mining is much higher than the benchmarks.   The 1080 and 1070 ti on my developer machine have a benchmark of 18 mh/s each for Tribus. Yet when I mine that algorithm on that pool, the miner's accepted hash rate returned by the miner is 77 mh/s for each card.  Even the pool is showing my hash rate around 60 mh/s for each card after only a couple minutes of mining.

Assuming your 1080 tis have a benchmark around 25 mh/s for tribus, the profit switcher was calculating your estimated earnings around $3.00, which falls beneath your price spike limit. However, the estimated earnings shown by the miner are based on the miner's current hash rate which is four times your benchmarked rate, causing the higher earnings estimate that you saw. However, the Price Spike Limit currently doesn't apply once you are actually mining.

To determine if the issue is with the pool or the miner, I started mining tribus on HashRefinery and got similar hash rates around 75 mh/s.  So the issue appears to be that Tpruvot miner benchmarks tribus at a third or less of what the devices can actually do.  If you are using the Update Benchmarks with Actual Rates feature, the software will revise the Tribus benchmarks automatically so that they reflect the higher real-world numbers. Then, if the estimated earnings for Tribus is higher than your price limit, it will be ignored by the profit switcher. Alternatively, you could manually update your device benchmarks to reflect those higher hash rates. In either case, once the benchmarks reflect the real hash rates, your Price Spike Limit will work for Tribus.

Graft looks to have xmr-64bit and xmrig support, so it seems possible no?

I've had similar experiences (8x 1080ti's). I'm still trying to find what pools I find preferable over others; no definitive answer yet.  I've had the same zergpool experience, though.

One I do like in particular (but can't prove earnings are any higher) is Blazepool.  Really responsive dev team on Bitcointalk; new coins added regularly, one of the better UI experiences (although they admit it is limited currently, expanding in future), and # miners are increasing almost daily.   Check out Blazepool.

giving ahashpool a try.  right off the bat it starts mining tribus.  Miners tab showing 94mhs and mining stats are showing $12.45.  I know this is wrong.  I have spike level set to $5.00 but it still mines ignoring my $5.00 limit.

I'll keep my eye on them, but my concern is that neither of these coins has a lot of pool or miner support right now.  Unfortunately, that makes them a bit of a challenge to incorporate into Hash Auger at this time. Hopefully that changes in the future as these coins gain traction.

Actually, Unimining.net is not an auto-exchange pool so it won't support bitcoin, litecoin or any coin that the pool isn't actively mining.  HashAuger actually filters out bitcoin wallets from being used with these pools so that people can't try to mine with them because these pools usually won't payout any earnings if you use the wrong type of wallet. I should probably do the same for Litecoin. Fortunately, many of these pools reject connections if you try to use an unsupported wallet. For the pools that don't auto-exchange, you must use a wallet for the coin you're trying to mine (Raven or Lux for example).

One of my plans for the next release is to better organize the pools so it is more apparent which support auto-exchange and which do not. In the meantime, the pools that support auto-exchange to bitcoin are: AHashPool, BlazePool, BlockMasters, HashRefinery, MiningPoolHub, NiceHash, Zergpool and Zpool.  

Tuning your devices will most likely improve your profits. It does take time, because you have to make small increments to boost performance without affecting stability. I suggested first finding a pool that you like because it should take less time and should increase your earnings more quickly. But you should definitely pursue both options.

All that is understandable.  Thanks.  But for my curiosity what can be expected if I really try to fine tune and really chase the profits with my cards? 
I am trying unimining.net and not having luck getting it going.  I assume it pays out in bitcoin.  I have my bitcoinr address from exodus.  I added a new wallet in HA  with my bitcoin wallet address.  In unimining.net setup under wallets I dont have a linked box that allows me to chose my new bitcoin wallet.  Only my old litecoin wallet is available.  If I try to start mining with litecoin selected the miner wont start. 
I did go to uniminer.net and added my exodus bitcoin wallet address but it does not want to stick. 

I think you could do a bit better with a Bitcoin wallet as it would give you the opportunity to try other pools such as AHashPool, BlazePool and HashRefinery. You might be able to do better still if you mined specific coins, but then you are betting that the value of those coin will rise more in the future - which is a risk that not everyone wants to take. Also, some people prefer the convenience of auto-exchange instead of managing multiple wallets and doing manual exchanges. It just depends on how involved you want to be and where you see the market headed.

Whether you decide to stick with Litecoin or try a Bitcoin wallet, you may want to try adjusting the Minimum Price Switch % in the software.  Right now, I've trying a 15% minimum to see if that is more profitable. There is a slight loss of productivity when you switch algorithms in any software because of the time it takes it shut down the current miner, start a new miner and then wait for the new miner to initialize and get work.  This process usually takes less than minute per switch, but the time can add up over the course of day. The default value of 10% covers this cost even if you are switching at the fastest interval (10 minutes). Increasing this value to 12 or 15% may improve profitability by preventing the software from switching algorithms when it is only marginally profitable to do so.

You can improve profitability by tweaking your devices with both overclocks and custom algorithm intensities. However, I think the biggest factor will always be pool prices. If you don't like the pool you're using, it would be better to try a different pool than spend too much time trying to boost your hash rate. An extra few percent increase in hash rates isn't going to make up for a pool that pays out a bit less than you are expecting.

Ok. I have had six straight days of running hashauger without any downtime except to apply updates.  I have 5 1080ti's.
I am mining zergpool and converting to litecoin.   I have hashauger on fully automatic.   I have banked $50 in the six days.  Should i be getting better and if so can we start a collective and talk about what is or is not working for you?   i would like to keep hashauger in the mix and try different setups that you are working with.   I would love to see at least $3.00 a day for my 1080ti but it is looking that the best I can do is about $1.80 on a several day average.  I am not opposed to converting to another coin.  I have not setup up a wallet yet to convert to fiat.  I really think hashauger is developing into a pretty great utility and I completely trust this app.  I really dont trust any of the pools I have used.  
So If your doing better with a 1080ti can you lend some assistance?  Or show me some proof to give motivation to do more tweaking.

Is it possible to add Graft and Verium coins please?
https://bitcointalksearch.org/topic/graft-coin-grf-cpugpu-mining-step-by-step-explained-tutorial-video-2956324
https://bitcointalksearch.org/topic/annvrm-verium-powt-cpu-mining-scrypt-scaling-new-roadmap-released-1540023

No problem. Thanks for the suggestion.

Hash Auger version 1.7.2 now uses the Cuda 9.1 version of Klaust; thanks for bringing that to my attention.  You shouldn't have to re-benchmark your devices, but a warning message will be displayed about the miner having been deleted as the file name changed. Also, the software doesn't yet delete old versions of the miners, so you may want to manually delete the previous version of Klaust if disk space is a concern. Also, since the upgrade replaces the Hash Auger .exe file, it may trigger a second developer contribution if one has already run for the day.  I'm currently working on a fix for this that I am planning to include in 1.8.