It Is All About BTC, LTC, ETH, DOGE, KAS mining as well as other alternative crypto currencies
The Rigel Nvidia GPU miner is a newcomer among the mining software solutions for GPU miners and more specifically for Nvidia GPU mining rig operators, but it is doing quite well in terms of competing with other older and more established and widely used miners. It does not support that many algorithms, but it is quickly introducing support for ones that are gaining a lot of user attention and the performance is really good. Not to mention that the miner comes with a nice text-based terminal user interface, is quite easy to use and comes with the right number of features including full overclocking set for the GPUs.
The Rigel miner also supports dual mining with Zilliqa (ZIL) as a means to further increase mining profitability as this dual-mining mode supports not only ethash and ethash, but any single or even dual algorithm combination + ZIL. Triple mining is supported for ethash + kheavyhash + zil and etchash + kheavyhash + zil, but that one is not actually that interesting of a mode compared to the dual-mining any algorithm with ZIL. The reason for that is due to the way ZIL is being mined – just a very short period of time every two hours, so essentially you retain the full hashrate of the main algorithm for mining the rest of the time and the extra profit from the mined ZIL is not something to miss.
In the last few versions ZIL mining was broken apparently, though the latest update of Rigel 1.3.4 brings it back working properly, so we are going to do a quick overview on how you can dual-mine Kaspa (KAS) with Zilliqa (ZIL), so that you can maximize your profit and get more than just single mining KAS. Now, Kaspa (KAS) and its kheavyhash algorithm is GPU-intensive, so you can optimize performance with a significant reduction in power usage compared to the normal settings for your video card. On the other end however you have Zilliqa (ZIL) that is a memory-intensive algorithm that just like Ethereum or any other Ethash-based crypto coin like ETC that is still mineable doesn’t need a lot of GPU power, but can benefit from maximum clock of the video memory.
So, how to combine these two seemingly opposite algorithms for dual-mining while getting optimal performance mining both? This is what we are going to take a look at now with the below example for dual-mining KAS + ZIL on an Nvidia GeForce RTX 3070 GPU, followed by explanation what and why is being used like that (the example is for windows, but the same settings should work on Linux as well). Make sure that in the example you set YOUR_KASPA_WALLET, YOUR_ZIL_WALLET and YOUR_WORKER_ID in order for the mining to properly start and you get the mined coins credited. We are using WoolyPooly as KAS mining pool and ShardPool for ZIL, though others should work as well (make sure you have the right settings for them):
rigel.exe -a kheavyhash+zil ^
-o [1]stratum+tcp://pool.woolypooly.com:3112 -u [1]YOUR_KASPA_WALLET ^
-o [2]zmp+tcp://eu1-zil.shardpool.io:3333 -u [2]YOUR_ZIL_WALLET ^
-w YOUR_WORKER_ID --log-file logs/miner.log ^
--cclock 300 ^
--lock-cclock [1]1710 --lock-mclock [1]807 ^
--lock-cclock [2]1200 --mclock [2]1000
Now, the cclock 300 option above sets the GPU clock offset in order for the video cards graphical processor to run at a lower voltage (it is set for all coins mined), do note that the 250-300 setting generally works well on most RTX 3070 GPUs, so test on your mining hardware and find what works stable for you. Since the first coin we are mining is KAS and we have it marked with [1] above then the lock-cclock [1]1710 and lock-mclock [1]807 options refer to the video card settings for Kaspa mining, setting the GPU clock locked at 1710 MHz and the memory clock to the minimum supported 807 MHz in order to reduce the power usage from the memory that we do not need much for the kheavyhash algorithm. The other line lock-cclock [2]1200 and mclock [2]1000 refers to the GPU settings for the second coin we are mining and in this case this is Zilliqa where we don’t need a higher clocked GPU, but could use a +1000 MHz overclock of the video memory over its stock settings in order to get the hashrate up to about 60 MH/s for an RTX 3070 GPU.
You will notice that mining with the settings used in the example above on RTX 3070 GPUs you will be getting around 90-95 Watts of power usage per GPU while mining KAS and when the ZIL switch occurs the different settings that will be applied will bring up the power usage to about 115-120 Watts (these can vary from GPU model to GPU model). So, there will be a slight increase of power usage for a couple of minutes in total on a daily basis and thus there will be not much of a difference in the overall power usage. While mining ZIL you will not be mining KAS, but again the switch is for a short period of time and the earnings for the mined ZIL should be able to compensate for the time you will not be mining KAS.
But why not leave the GPU settings for KAS mining for ZIL mining as well some of you may ask? Well, the answer is pretty simple – the ZIL mining performance will be very low due to the low memory setting we are using to save power when mining KAS. While Kaspa can be mined without performance loss at 807 MHz for the video memory, that operating frequency will result in just around 5 MH/s of hashrate per RTX 3070 GPU compared to round 60 MH/s when the memory is overclocked with 1000 extra MHz over its stock frequency on the same GPU. So, while you might be saving a little bit of power the hashrate will be so low that you might not be able to send even one share during the ZIL mining timeframe and thus you might just be wasting your time, unlike what the GPU can do at 60 MH/s. Just as a reference, in a single ZIL mining period (every two hours) a 6x RTX 3070 GPU mining rig should be able to currently mine 1-2 ZIL (depending on the number of shares you get).
– To download the latest Rigel 1.3.4 Nvidia GPU miner with ZIL dual-mining support…
There are a lot of GPU-intensive crypto coins out there that can be mined with video cards that unlike ETH/ETC and other memory-intensive algorithms do not require a lot of gigabytes of video memory or a very fast memory access speeds or clocks. One interesting such projects is Dynex (DNX) and we’ve already covered it last month as something interesting that might deserve your attention and mining power (still Nvidia GPU mining only). Now, we are going to be focusing on a different aspect of mining DNX that uses its own custom miner that unlike most other multi-mining software does not support GPU tweaking options. This means that if you do not optimize the GPUs you are using for mining you will be wasting a lot of extra power without any gain in performance and at times of lower profitability for mining this is not something you should be eager to do.
Miners that use Linux-based operating systems such as HiveOS have more options to control the operating parameters of their GPU-based mining rigs regardless of what miner software they currently employ, however Windows users are having a hard time doing the same thing. There are some useful graphical tools such as MSI AfterBurner or command-line tools such as nvidia-smi, but they are not as useful or easy to use or functional as one might think. The good news here is that Windows users might use some “cheats” in easily tweaking their mining hardware the way they want it to, regardless of the miner software they utilize.
This is possible all thanks to some of the new features introduced lately in the lolMiner mining software. These are not only the options for command line setting the GPU core clock, memory clock, power limit and the core clock offset, but also and very importantly the option to turn off the reset of overclock settings when exiting the miner. This means that you can run the lolMiner software briefly (make sure it is ran as Administrator for clock settings to work), quit it and have it leave the settings you applied for the GPU clock and then just run another miner like DynexSolve for mining DNX or just about another one. Just make sure that the mining software you run after that does not manage GPU settings on its own or does reset them to some default states as it will defeat the whole purpose.
Below you can find an example command line to run lolMiner setting the clocks on RTX 3070 GPU and then automatically exiting without resetting them back to the standard ones when exiting. Since we run lolMiner without a real KAS address it connects to the pool and then automatically exits the miner and then we run the DynexSolve miner and start mining DNX with the same optimized GPU setting that we would’ve used for mining Kaspa for instance with lolMiner. This way we get the same or maybe even higher performance with lowered power usage than if we just run the DynexSolver miner with the default settings for the GPU, just don’t forget to add your DNX wallet in the example below (do not add KAS wallet, leave the x there):
lolMiner.exe --algo KASPA --pool stratum+tcp://pool.eu.woolypooly.com --port 3112 --user x --watchdog exit --cclk 1710 --mclk 810 --coff 300 --no-oc-reset
dynexsolvevs.exe -mining-address YOUR_WALLET -no-cpu -multi-gpu -stratum-url dynex.neuropool.net -stratum-port 19331 -stratum-password YOUR_WORKER_ID -stratum-paymentid YOUR_PAYMENT_ID
The same thing can be applied to another miner and another crypto coin that is GPU-intensive and you can lower the video memory to the minimum and also decrease the operating frequency of the GPU to a lower level that it can still handle mining at with a high-enough operating frequency. This way you can go significantly lower than 100 Watts per RTX 3070 GPU and still managing to maintain the pretty much same hashrate that you would normally achieve mining at stock settings. And if the coins tolerate even higher GPU clock and that brings extra performance you can utilize the extra power usage headroom for further increasing the performance you get while maintaining much lower power usage in general.
Radiant (RXD) is another project that is quickly catching up with crypto miners, so we are going to be sharing some tips on how you can optimize Nvidia GPUs for better performance with lower power usage for mining RXD with the Bzminer software. Radiant uses the SHA512256d algorithm, another GPU-intensive algorithm, so you can go as low as possible with the video memory and higher on the GPU clock to get better performance with lower power usage for this one. AMD GPU miners interested in mining Radiant (RXD) should go for the SRBMiner-MULTI CPU & AMD GPU Miner 1.1.3 software, however it does not come with so advanced GPU clocking features, so you might need to use different tools in trying to optimize the hardware. Here we’ll be focusing on Nvidia GPUs only and Bzminer.
We are going to be using the latest Bzminer v12.1.1 as mining software for the examples below and cover some of the popular Nvidia GPUs fo rmining with settings and expected hashrate and power usage. Obviously, we are going to be setting the GPU clock to a higher fixed frequency (oc_lock_core_clock) and the memory clock (oc_lock_memory_clock) to the lowest possible fixed frequency and we’ll add in the core offset (oc_core_clock_offset) to further lower the power usage by lowering the voltage of the core. In the examples below do not forget to change the pool if you want and also set your Radian wallet address in the place of the YOUR_RXD_WALLET, also do not forget that you need to run Bzminer as administrator in order to for the OC settings to function!
Nvidia GeForce RTX 3090 GPU running at 1.53 GH/s hashrate with around 170W of power usage:
bzminer -a radiant -p stratum+tcp://stratum-eu.rplant.xyz:7086 -w YOUR_RXD_WALLET --oc_lock_core_clock 1800 --oc_lock_memory_clock 807 --oc_core_clock_offset 350 --nc 1
Nvidia GeForce RTX 3090 GPU running at 1.21 GH/s hashrate with around 130W of power usage:
bzminer -a radiant -p stratum+tcp://stratum-eu.rplant.xyz:7086 -w YOUR_RXD_WALLET --oc_lock_core_clock 1410 --oc_lock_memory_clock 807 --oc_core_clock_offset 350 --nc 1
Nvidia GeForce RTX 3080 GPU running at 1.23 GH/s hashrate with around 145W of power usage:
bzminer -a radiant -p stratum+tcp://stratum-eu.rplant.xyz:7086 -w YOUR_RXD_WALLET --oc_lock_core_clock 1750 --oc_lock_memory_clock 810 --oc_core_clock_offset 250 --nc 1
Nvidia GeForce RTX 3080 GPU running at 1.00 GH/s hashrate with around 115W of power usage:
bzminer -a radiant -p stratum+tcp://stratum-eu.rplant.xyz:7086 -w YOUR_RXD_WALLET --oc_lock_core_clock 1425 --oc_lock_memory_clock 810 --oc_core_clock_offset 250 --nc 1
Nvidia GeForce RTX 3070 Ti GPU running at 0.85 GH/s hashrate with around 130W of power usage:
bzminer -a radiant -p stratum+tcp://stratum-eu.rplant.xyz:7086 -w 1YOUR_RXD_WALLET --oc_lock_core_clock 1710 --oc_lock_memory_clock 810 --oc_core_clock_offset 350 --nc 1
Nvidia GeForce RTX 3070 GPU running at 0.815 GH/s hashrate with around 90W of power usage:
bzminer -a radiant -p stratum+tcp://stratum-eu.rplant.xyz:7086 -w YOUR_RXD_WALLET --oc_lock_core_clock 1710 --oc_lock_memory_clock 810 --oc_core_clock_offset 350 --nc 1
Nvidia GeForce GTX 1080 Ti GPU running at 0.565 GH/s hashrate with around 100W of power usage:
bzminer -a radiant -p stratum+tcp://stratum-eu.rplant.xyz:7086 -w YOUR_RXD_WALLET --oc_lock_core_clock 1550 --oc_lock_memory_clock 810 --oc_core_clock_offset 250 --nc 1
Some more useful tips for anyone that wants to experiment with these settings above. We have provided good starting point in terms of frequency and offset that should be stable on most GPUs, however depending on your specific hardware and conditions you might be able to go up a bit for extra performance or power savings or you might need to go lower for attaining good stability. If Bzminer crashes or restarts you might want to start by lowering the setting of the oc_core_clock_offset parameter, that would lead to slightly higher power usage. If lowering the offset does not help, then keep its value and try lowering the GPU clock a bit. The video memory setting is already at the minimum level, so there is no need to play with the clocks there.