Posts Tagged ‘Grin AMD miner

The latest update of the AMD GPU miner lolMiner 0.8 brings support for the Cuckatoo 31 algorithm used by Grin, for GPUs with either 4GB and 8GB video memory (lower memory mens less performance) for Windows and Linux. To mine with the new algorithm you need to use the parameter --coin GRIN-AT31, NiceHash does not currently seem to be supported for C31 though. It is interesting to note that the Grin cycle finding is completely done on the GPU, so the miner has almost zero CPU load (on AMD, Nvidia has some due to OpenCL back end). Aside from Cuckatoo 31 support the new version also has some Stratum bug fixes (NiceHash) for Equihash 144/5, 192/7 and 96/5, improved general stability of the miner as well as added distinct 1GB, 3GB and 4GB as well as 6GB and 8GB kernel for mining MNX for Windows (the Linux version that already implemented had this in 0.7.1).

We remind you that lolMiner is a closed source OpenCL GPU miner available for Windows and Linux as pre-compiled binaries only and that there is a developer fee for using the software (1% for all supported algorithms). It might work on Nvidia with OpenCL, however stability and performance might not be optimal, so it is best for use with AMD GPUs.

To download and try the latest lolMiner 0.8 OpenCL miner for Windows or Linux…

Grin is a new crypto project for creating a the private and lightweight mimblewimble blockchain that would provide electronic transactions for everyone without censorship or restrictions. The algorithm that Grin uses is Cuckoo Cycle with 1 minute block time, 60 GRIN block reward and no limit on coin supply or reward halving. There is not going to be any ICO, founders reward, pre-mine or masternodes available for this project, so people that do not like any of these should be Ok with this project. Miners for Grin are still currently being developed, although there are some already announced or available, though initially it could be a bit of a challenge to make things work with the right miner and pool. For instance the official MWGrinPool is currently not taking new registrations, though new pools should be available as alternatives. There are already a couple of options for pools on the Testnet where you can test the mining before the Mainnet launch like Spark Pool.

The official open-source Grin Miner software is currently only available for Linux and Mac OS X and it supports both CUDA and OpenCL mining. For miners looking for Windows miners there is the Grin Gold Miner (with 1% fee for the Grin Development Fund and 1% fee for further miner development), although it also works with Linux) that supports all AMD and NVIDIA 8GB cards (requires a GPU with 8GB VRAM!). Do note however that Nvdia RTX 8GB cards need Linux or Windows 7 due to memory addressing issues. RTX 2080 and RTX 2080 Ti are fastest in terms of performance if you manage to make them work, OpenCL code for AMD cards is currently still being optimized. Cheaper slower Celeron CPUs are not recommended for many cards on the system (similar to Merit and BITC that also use the same algorithm for mining). Another alternative for Windows miners should be the GrinPro Miner that is not yet available for download and that will have 2% developer fee. This miner should work on AMD and Nvidia GPUs with 8GB VRAM, although it promises upcoming support for 6GB VRAM cards as well, so that could be interesting for some users.

Unfortunately zjazz, who was pretty active in miner development for Cuckoo Cycle support for Merit and BITC hasn’t updated his miner for a while and will probably not be supporting Grin mining at least for the moment. The latest version of Bminer that was just made available however also comes with experimental support for mining Grin, so you might want to check it out as well. Do note that it is a closed source miner with 2% developer fee, so have that in mind as well when checking out different miners for Grin. The latest version of the closed source miner NBMiner has also just added support for the Cuckaroo algorithm used by Grin (both Windows and Linux), so you can give it a try as well (Nvidia only).


top