CPUMINER-GR – CPUMINER-GR – Optimized version of GR miner for RTM
cpuminer-gr is a fork of cpuminer-opt by Jay D Dee which is a fork of cpuminer-multi with optimizations imported from other miners developed by lucas Jones, djm34, Wolf0, pooler, Jeff garzik, ig0tik3d, elmad, palmd and Optiminer. , with additional optimization by Jay D Dee. All code is considered open and free. Cpuminer-opt is a CPU mining software. There are two versions in the archive. Cpuminer-opt and Cpuminer Opts for VeriumReserve cryptocurrency. The version cpuminer-gr is a fork of cpuminer-opt by JD, which is a fork of cpuminer-multi with optimizations imported from other miners.
Miners are often flagged as malware by antivirus software. This is a false positive, they are flagged simply because they are cryptocurrency miners. The source code is open to the public . If you don’t trust the software, don’t use it. See the RELEASE_NOTES file for the changelog and INSTALL_LINUX or INSTALL_WINDOWS for compilation instructions.
X86-64 architecture processor with minimal SSE2 support. This includes Intel Core2 and newer models and AMD equivalents. Additional optimizations are available for some algorithms for processors with AES, AVX, AVX2, SHA , AVX512 and VAES.
ARM and Aarch64 processors are not yet supported.
64-bit Linux or Windows OS. Ubuntu and Fedora-based distributions, including Mint and Centos, are known to work and have all dependencies in their repositories. Others may work , but may require more effort. Older versions such as Centos 6 don’t work due to lack of features. Windows 64-bit is supported with mingw-w64 and msys or prebuilt binaries.
MacOS, OSx and Android are not supported.
A Stratum pool that supports stratum+tcp:// or stratum+ssl:// protocols or RPC operation using http:// or https://. GBT is YMMV .
CPUMINER Quick Setup -GR
To add or use miner options, use the attached config.json file. All options must be in JSON format, for example: “long-flag-name”: “Some_value”.
A full list of miner parameters and other tips can be found in the readme.txt file.
Tuning starts automatically when you start the miner. If a previous tune tune_config file exists (or --tune-config=FILEflag is used), it is used instead. This behavior can be overridden with --no-tuneor --force-tune. On non-AVX2 processors, the process the default setup takes about 69 minutes. On AVX2 processors, the default setup process takes about 155 minutes.
A little explanation of what tuning does. The traditional way of hashing would be to take some input, hash it to create an output hash. This is what might be called normal hashing (aka 1way) since we do 1 hash at a time. What we can do is hash 2 or 4 hashes at the same time! Due to the different options used in each block, etc., memory requirements change and we would like to have as much as possible in cache as RAM is SLOW! So 1way would require 128KB, 256KB, 256KB, 512KB, 1MB or 2MB to store somewhere if we want to solve this problem. 2way will need twice that amount and 4way 4x. In some cases, such as the 256KB variants, this increases the requirement to around 1MB, or up to 2MB for the 512KB variants. In most cases, this amount of data can fit in the cache, which can improve performance. Too high a value can If you can’t reduce it, imagine 8MB for 4way Fast (2MB option) (it’s still faster in some cases if your CPU lacks cache, e.g. i3 or some mobile CPUs).
So there are 6 variations that make 20 possible “rotations”. And we test all of these 20 rotations to see if we have brought a one-way, two-way or four-way movement in each of them an improvement or not. We can’t check individually when they hash using all of them, it might not be accurate anymore, so we have 8 scenarios per rotation for AVX (only 1 or 2, 2^3, 2 solutions for 3 choices) and 27 for AVX2 (1way, 2way, 4way, 3^3, 3 solutions for 3 options) -> it’s –tune-full which checks everything. –tune-simple only checks 4way on the Turtle and Turtlelite variants, and by default also checks the Dark and Darklite variants as they are most likely to be used or benefit processors during in all of our tests. This way we can use the most cache and use it as efficiently as possible. The question you might ask is how to do 2way faster than doing only 1way 2 times? This is because we We can also use parallelization and other tricks to notify the CPU of what’s going on so it can prepare the data faster.
Ghost Rider (GR) algorithm
The Ghost Rider (GR) algorithm used for RTM mining consists of 15 “basic” algorithms (same as X16 without SHA) and 6 different Cryptonight variants, of which only 3 are used for hashing. Each block (actually part of the previous block) dictates in which order these algorithms should be calculated and which of the Cryptonight variants should be used. It looks like this: 5 main algorithms, 1 Cryptonight, 5 cores, 1 Cryptonight, 5 cores, 1 Cryptonight. ;
As you can see, all the main algorithms are always used, but only 3 of the 6 Cryptonight variants. These parts of Cryptonight are the slowest/most complex part of the whole hashing process. The kernel algorithms work very well on almost every processor, but Cryptonight requires a certain amount of memory if that memory can be stored entirely or mostly in cache. this will greatly improve performance (the main reason AMD Ryzen is much faster than Intel as they have much more L3 cache). ;CN variants use 128KB, 256KB, 256KB, 512KB, 1MB, or 2MB, or memory. Also, the more memory required, the more iterations through it (making it even slower). Performance these algorithms can vary greatly. You can get around 1500 Hz/s if the blocks need the 3 slowest options (the ones that use the most memory), and 10000 + H/s if the blocks use the most fast (on the example of my 2x2698v3).
This means 2 things. Firstly, the hash rate, as you have noticed, is very volatile and almost always changes with each block. The miner should show you the Cryptonight variants used in the current block, Turtlelite, Turtle, Darklite , Dark, Lite, Fast are options in the same order as Memory, the irony is that Fast is the slowest option and Turtlelite is the fastest :). Secondly, some blocks (in most cases those that use simpler variants) are faster and not very consistent. In the long run, everything should average around 120 seconds per block.
Users are encouraged to post their bug reports using git Issues or on the official RTM Discord, or by opening an issue in git:
All problem reports should be accompanied by a proper problem definition. This should include how the problem occurred, the command line, and miner output showing startup messages and any errors. History is also helpful, i.e. did it work before.
Generate wallet and RTM address
Step 1: Download and install the Raptoreum Core wallet
Step 2: Synchronization of the Raptoreum blockchain
Once you have installed the wallet, it will first be out of sync with the network – it will take a little time to sync – you can see the progress in the bottom left corner of the wallet window. Just go watch a series of anime that you like so much and you’ll be done.
Step 3: Create an address to accept coins
When your wallet is synced with the network, click “File” in the upper left corner of the wallet window, and then click “Get Addresses” as shown in the image below. Click “Create”, give your wallet a name, and then click “OK”.
You now have a Raptoreum wallet address! In the get addresses window you should now see your address, which if you right click you can copy it to paste it somewhere else.
Start mining Raptoreum
First, download the official wallet to get the address where our good will be mined. Download from here https://raptoreum.com/ via the downloads tab.
After installing and synchronizing the wallet (synchronization is optional to get the wallet address). We go to the wallet – file – Receipt address & nbsp; (here we already have a wallet, but you can create one or more) – copy the wallet address.
Select the Pool. I really liked the pool https://flockpool.com/ (convenient, not buggy , everything normally reflects, adequate in one word). There are others, but this one is better for me, and we will talk about it. We go to the pool website and click Start Mining. Then we select our region (check Europe), paste the previously copied RTM wallet address and set the name workera (x for example), I did not set a password.
Next, download and install CPUminer-gr. Go to github, download the version of the miner we need.
Unpack. Change the batch file called config, mine looks like this:
where “user” should be your wallet address, where “threads” is the number of threads involved in mining. “0” – use all threads (in my case 11 out of 12)
Then run the miner itself (cpuminer) as administrator. At the first start, the tune_config file should be generated in the miner folder (it takes 2-3 hours), the next time you start it, you won’t have to wait. After generating tune_config, your miner will start showing up on the pool.
Payouts on the pool are automatic, twice a day (morning and night). Everything comes without problems, stable. That’s all on this, I attach my result for 5 days:
Remember that you can reuse the tune_configand config.json
files from version 1.2.X by copying them to the main folder 126.96.36.199.
Users of Alder Lake (Intel 12th generation) are strongly advised to reconfigure the miner.
Make sure the number of cores E has been determined correctly!
Fixed issue with miner reconnecting after donating to the same pool. Happened if user switched to backup pool at least once.
Fix issue with large pages, AVX or lower when tune-fullis set to true.
HOTFIX: Fix fallback layer issue. May cause to the fact that the miner will not start properly.
General update recommendations:
Update your startup scripts to reflect the changes in Windows 11.
Use the new avx2-sha-vaes on Alder Lake.
Fix some output inconsistencies during setup.
Added improvements to Alder Lake tuning. ~6% (12900k) hashrate increase. REQUIRED system retuning to improve performance.
Add --ecores=N– Specify the number of cores E of the processor. The miner will try to autodetect. It will show the predicted number. Incorrect or on a platform that does NOT support E-cores (Intel Alder Lake) may result in degraded performance.
Add --disable-rot=LIST—separated comma list of spins that will not be used when mining. For testing purposes only! Using this for normal mining will decrease your overall average hashrate!! Does not apply to the setup process.
When you get rich, you can support the author in RTM: RUZb2pp45x5qAjbS3usXAGW8BzK1fvKJBo
Do you like to read similar articles about cryptocurrencies and mining, do you want to support the author or ask questions? Be the first to know the news, subscribe to my telegram channel: t.me/cryptomineg