Do HDD specs matter at all?

  • Hey guys,

    I was wondering whether the RPM, cache size, etc. matter besides the amount of space plotted on the HDD?

    For example, here is a section from my miner right now:

    19:27:27 New block 250328, baseTarget 2272889, netDiff 8062 Tb
    19:27:31 Thread "F:\Burst\plots" @ 3.9 sec (26.5 MB/s) CPU 34.62%
    19:27:32 [17737134402712869425] found DL: 337793
    19:27:32 [17737134402712869425] sent DL: 337793 3d 21:49:53
    19:27:33 Thread "I:\Burst\plots" @ 6.2 sec (16.4 MB/s) CPU 25.71%
    19:27:48 [17737134402712869425] confirmed DL: 337793 3d 21:49:53

    Both plots are 409 gigs, but F:/ is a 32 MB cache drive running on SATA III, while I:/ is a 16 MB cache external drive running on USB 3.0. As you see, I:/ runs around 10 MB/s slower than F:/.
    Is this difference caused by the fact that I:/ runs on USB, or is it the result of the 16 MB cache?


  • This is a by product of the USB bus. To make you feel better I have an external drive on a USB (old 1.1?) that is mining and plotting at the same time and it gets a read through of about .4mb/sec; Yes that is point 4. Takes about 3 minutes to read 600GB, once the plotting is done speeds increase to about 5mb/s and then there's no problems. But if your not plotting your not growing.


  • Thanks for the intel, IceBurst. I think I will try and put that external drive on SATA to see if there still remains a difference. I'll post the result here soon.

  • Hey,

    I'm back with some results.

    19:51:37 New block 250693, baseTarget 1709610, netDiff 10718 Tb
    19:51:38 [17737134402712869425] found DL: 3468572
    19:51:38 [17737134402712869425] sent DL: 3468572 40d 03:29:32
    19:51:39 [ 7397608791598125259] found DL: 361141
    19:51:39 [ 7397608791598125259] sent DL: 361141 4d 04:19:01
    19:51:40 Thread "F:\Burst\plots" @ 3.9 sec (26.4 MB/s) CPU 31.36%
    19:51:41 Thread "I:\Burst\plots" @ 4.6 sec (22.2 MB/s) CPU 26.92%
    19:52:18 [17737134402712869425] confirmed DL: 3468572 40d 03:29:32
    19:52:21 [ 7397608791598125259] confirmed DL: 361141 4d 04:19:01

    I:\ gives 3-4 MB/s lower rates than F:. Since this shows a significant improvement compared to what I:\ was producing through USB, the remaining difference (or a major part of it at least) derives from the fact that I:\ is SATA II compatible, whereas F:\ is SATA III compatible (both of them are plugged into SATA III slots on the motherboard).

    I cannot clearly state the relevance of RPM or cache size to the actual mining capabilities of the drives, but, as suggested by IceBurst, the type of connection used seems to be a major factor.

    Anyone else on the RPM and cache thing?

  • try a USB 3.0 lets see if it would improve

  • @limpopo my hdd case is USB 2.0 compatible, but once I got ahold of a 3.0 compatible case, I'll post my test results here