CPU vs GPU mining

  • Hi, I am new here! I am wondering, if someone has the same amount of HD capacity (e.g 10tb) will he generate more burst coins with GPU mining or the difference in the reading speed does not affect that (assuming that all the HDs are connected in the same PC).

  • admin

    @b11001h19 Reading speed is important, but there is unlikely to be much difference in CPU vs GP on 10TB.

  • @haitch Thank you for your reply! When the difference starts becoming considerable? 100tb?

  • admin

    @b11001h19 It depends. Number of drives, CPU and GPU capabilities. When the CPU becomes the bottleneck, GPU may be faster.

  • As haitch said it depends on the Hardware you are running and without knowing the concrete specs of the Hardware you are running its virtually impossible to answer that question.

    As a rule of thumb for "only" 10TB of mining (2 Hdds or three) its not worth hassling around with gpu mining (especially on an older machine..) instead of cpu because the differences are negligible on an average home pc bought like 3-5 years ago.

    Technically there is no doubt that there IS an Advantage of GPU over CPU (assuming you dont hold an technically outaged Laptop with internal graphic Card which prob. would cause all other Kind of (Performance) Problems.

    But talking about difference the net difference @ 10Tb would be Close to None, talking about 1-5$ / month if ever. Depending on your Hardware as mentioned above.

  • @haitch , @Marc thank you both for your responses, with bottleneck it means that the reading time is more than the time between blocks?

  • @b11001h19 No.

    Bottleneck in this term refers to some mis/bad Balance of the Hardware you are using and/or the one component which is responsible for this., aka the weakest link in the whole chain.

    Imagine having a machine with the newest CPU worth $1.000, a Graphic Card worth another $1.000 , 64GB of RAM, SSD etc. etc.. - but for some strange reason connect to that Beast an external HD to it which only uses USB 2.0.

    USB 2.0 here would be the "bottleneck" when in Comes to read/write speeed - as your machine could deliver much more more, but the Connection type of USB 2.0 just doesnt allow more throughput, even if you put Googles Server Farms prozessing powers on the other end.

    The other way round, if your components technically are able to deliver x (SATA Ports, USB ports, Hubs, Raid Controllers etc.) but your cpu/gpu is too weak and only able to deliver 1/2 x to make up for that whats possible the cpu/gpu becomes the "bottleneck". IF you are looking for higher numbers of TB, practically starting at 50-100 TB afaik (For 10 TB in practical life for plotting/mining you can ignore that).

    Talking about a Porsche 650 horsepower but only equipped with 125 wheels - or the other way round.

  • @Marc very nice explanation!