Posts Tagged ‘ccminer maxwell optimized


We have compiled a new windows binary of the ccMiner 1.4.7.SP (source) fork with optimized kernels for Nvidia Maxwell GPUs supporting Compute 5.0 (GTX 750 / 750 Ti) and Compute 5.2 (GTX 970 / GTX 980) that includes some slight performance improvements for the Blake, Echo and Shavite algorithms as well as other more general optimizations that should result in slightly faster hashrate on some of the algorithms supported by the miner. The windows binary you can download below is compiled with VS2013 and CUDA 6.5 and is only for Nvidia Maxwell GPUs, it will not worl on older hardware – for older Nvidia cards you can use tpruvot’s ccMiner fork that should also work on Compute 3.0 and 3.5 GPUs.

You can download the ccMiner fork by SP optimized for Maxwell GPUs for Windows OS here…


There is a new beta version of the ccMiner fork optimized especially for miners using Maxwell-based GPUs from Nvidia such as GTX 750, GTX 750 Ti, GTX 970 and GTX 980 supporting Compute 5.0 or 5.2 by SP. The new ccMiner release 1.4.7.SP (source) comes with some performance improvements for GTX 980 for the X11, X13, X15 algorithms as well as with multiple improvements for other algorithm supported and some increase in hashrate for GTX 750 Ti as well. The initial 1.4.7 release did have some issues with the X-based algorithms on GTX 750 however, so we’ve waited for these to be fixed and now it should work on GTX 750 Ti as well. We have a windows binary available from the latest source with all of the fixes and improvements available for download, it is compiled with VS2013 and the latest CUDA 6.5 version available in a single binary with support for both Compute 5.0 and Compute 5.2. Do note that older Nvidia video cards are not supported with this release, you should go for the latest tpruvot miner for older Nvidia-based GPUs. On a non-overclocked GeForce GTX 980 we are now getting up to about 7700 KHS for X11 as opposed to about 7550 KHS using the previous 1.4.6.SP beta miner that we have posted a windows binary of a few days ago, on GTX 750 Ti the improvements are not so notable.

You can download the ccMiner fork by SP optimized for Maxwell GPUs for Windows OS here…


This is something that will be of interest for Nvidia GPU miners that use GeForce GTX 750/750 Ti or the newer GTX 970/980 GPUs – a new ccMiner fork with optimized kernels for the Maxwell architecture (Compute 5.0 or 5.2) by SP. The ccMiner release 1.4.6.sp.Beta6 (source) is forked from tpruvot’s release supporting most of the available algorithms, but the focus here is to get the maximum performance out of the Maxwell GPUs. We have compiled a windows binary of the miner that you can download from the below, it is a 32-bit version compiled with VS2013 and with the latest CUDA 6.5 that has official support for GTX 970 and GTX 980 GPUs. Do note that archive includes a compiled binary for Compute 5.0, so it will not work on older video cards that are not based on the Maxwell architecture. Our tests have show that on a stock GeForce GTX 980 video card with no overclock we are getting about 7550 KHS for X11 up from about 6500 with the ccMiner 1.4.5-tpruvot under the same conditions and on the same hardware, X13 is up to about 6000 KHS with the SP optimized miner versus about 5300 KHS with the version from tpruvot. So definitely worth trying if you have some Maxwell hardware and you are mining with it to get the best possible performance, overclocking of course also brings the hashrate higher.

You can download the ccMiner fork by SP optimized for Maxwell GPUs for Windows OS here…