Posts Tagged ‘SIA pool mining

sia-gpu-miner-genoil

Siacoin (SC) is getting a lot of attention lately and there is now a new fork of the miner available made by Genoil (source). The Sia GPU Miner fork by Genoil is for OpenCL, though it works on both AMD and Nvidia and provides some performance optimizations along with other interesting features such as the ability to load external OpenCL kernel file that might be useful for people that want to try modifying or optimizing the CL file. Genoil reports that his optimized OpenCL kernel is capable of providing roughly +18% performance increase on a 7950 and we are seeing some performance boost on Nvidia as well, though more like 3-4% boost on a GTX 970. Nevertheless the fork by Genoil does offer some performance improvement over the Siacoin default miner as it has taken that OpenCL Go miner kernel as a base and even more improvements could be possible for faster performance.

We have a working 64-bit Windows binary available for download below that you might want to try. We have tested it and can confirm that it works on both Sia Nanopool and Siamining pool, though for some reason the hashrate reported by the miner is a bit higher for Nanopool than for the Siamining pool. Do note that this miner does require you to run a separate instance for each GPU in your system, it will by default run only on the first video card it finds. So while things do look very promising for this fork it apparently can use a bit more work to become as easy to be used as the Siacoin Go miner and maybe even better performance wise.

To download and try the Sia GPU Miner for OpenCL Forked by Genoil for Windows…

sia-cuda-miner

Sia CUDA Miner is a GPU miner designed for mining Siacoins on Nvidia GPUs using CUDA instead of OpenCL like on the Siacoin Go pool miner. The Sia CUDA Miner is being developed by KlausT (source) and works pretty well already, though it still needs some more work in terms of available features and in terms of performance optimizations. Our tests have shown that using the Sia CUDA Miner on various Nvidia GPUs does result in a hashrate that is a bit lower than what you can get using the OpenCL miner on the same hardware. Another limitation that the CUDA miner has is that by default it runs on a single GPU and you will need multiple instances running to have it work on multiple GPUs. Also, although the miner supports pool mining it apparently only works with pools that have long pooling enabled and this means that for the moment you are limited to the Siamining pool.

We have compiled a 64-bit Windows binary using CUDA 8.0 available for download below, our tests have shown that this was the fastest performing versions after trying different versions of CUDA and 32-bit compilation. The Windows binary is based on the latest source code and is version 4.01 that should provide more stable and problem free pool mining as compared to earlier releases. You are welcome to try the miner and report your results, but as we’ve already said the OpenCL miner is still the better choice for Nvidia GPUs at the moment in terms of features and support as well as performance wise in terms of hashrate, so you might want to stick to the Sia Go Pool miner that uses OpenCL.

To download and try the Sia CUDA Miner 4.01 for Nvidia GPUs on Windows…

siamining-siacoin-mining-pool

A few days ago we’ve covered the beta SIA pool from Nanopool, and now there is a second more interesting SIacoin mining pool open for public already available called Siamining. Nanopool is still in beta and apparently has some delays with payments, along with high minimum limits and uses PPLNS payment system, while the Siamining pool does seem to be out of beta already and offers more interesting terms, especially for smaller miners. The Siamining Siacoin pool uses a PPS (Pay Per Share) payment system with a slightly higher fee of 3% (to cover for the higher risk for the pool) and does payments every 6 hours. So no need to wait for a very high minimum balance (2000 SC minimum) or to wait for block confirmations before you get your SC payments sent to you, there is also a Long Polling support to minimize stale shares. We have been testing the pool for about a day already and it does seem to be working quite well already, so we do recommend that you check it out, especially for smaller miners.

Meanwhile we have also done some benchmarks using the Siacoin Go Pool miner (source) that uses OpenCL for mining SC on different AMD and Nvidia GPUs (the miner supports mining on Nvidia using OpenCL) and you can fidn the results below. Do note that Nvidia is generally slower than AMD in terms of performance with OpenCL, however the new Pascal Nvidia GPUs such as GTX 1070 and GTX 1080 manage to do quite well in terms of hashrate. When you factor in the very good profitability mining Siacoin it is worth checking it out, especially if you are wondering shoudl you continue mining Ethereum with the current low exchange rate and high difficulty when there are already some good alternatives such as SIA.

Siacoin Go Pool miner hashrate:
– GTX 1080 – 1945 MHS
– GTX 1070 – 1466 MHS
– GTX 980 Ti – 1220 MHS
– GTX 970 – 803 MHS
– GTX 950 – 385 MHS
– GTX 750 Ti – 301 MHS
– RX 480 – 872 MHS
– R9 280X – 849 MHS
– R9 290x – 1116 MHS

Alternatively you can go for dual mining Ethereum and Siacoin using the latest Claymore Dual Miner with ETH in the main focus and SC as a secondary crypto coin to fully utilize the resources of the GPU you are using. Just a reminder that the dual miner is only for AMD GPUs, so for Nvidia GPUs you still need to go for the SIA Go miner that you can download below (64-bit binary for Windows). A word of warning, the Siacoin miner is pretty heavy for the GPU, so be careful with the temperatures and power usage if your focus was lately on mining Ethereum that does not stress the GPU that much unlike other algorithms. The Siacoin Go Pool miner easily hits the TDP limits of the GPU while mining, so make sure your video cards are cool and that your power supply can handle the load.

To download and try the Siacoin Go pool miner for AMD and Nvidia GPUs…


top