Based on your description of the HDTune throughput graph (I guess that's what you mean when you said "timeline trail"), then the Samsung drive likely has problems, as seed.helper said.

Look at the full, raw, SMART data from the Samsung drive (i.e. the full set of values where you saw attribute 0xC7 = 1), and check the number of reallocated sectors, and the number of sectors pending reallocation. You'll see more info on the Wikipedia page on SMART that I referred you to, on your other thread. Significant values in either of those two attributes is bad news. However your ability to RMA a drive may be disputed by some suppliers, until an actual SMART failure is reported.

FYI in future it would help if you supply the actual info which you are looking at (e.g. SMART data, throughput graphs etc.) instead of trying to describe everything in words. For example, just because you think that HDSentinal Pro showed no problems, does not mean that others here would agree, if you showed the actual data e.g. the raw SMART data and graphs from HDTune

the HDtune graph is very erratic, the timeline trail nosedives all over the place. My lowest read speed is about 19/ms whereas my highest is 147mbs.

Thanks for that data.

Originally you said:


The HDTune graph you just supplied does not show "nosedives all over the place" - at least based on my interpretation of what you mean, but we haven't seen that original graph, of course. There's a small dip at around 25GB, but that's the only slight concern I have. The lowest read speed of 73.6 MB/s (at ID) is about half the max read speed of 147.5 MB/s (at OD), as expected.

The SMART data suggests there are no sectors reallocated or pending reallocation, and no other obvious areas of concern.

So in summary, in the data you've just supplied, I see no clear evidence of a problem with that disk drive. If you get different HDTune results, but only when booted from that Samsung disk (assuming all hardware connections stay the same for all tests), I suggest that perhaps the benchmark is being affected by background programs accessing that disk at the same time as the benchmark is running.

I would run the HDTune read benchmark several times in the current configuration, so that you can build confidence that the good result is not just a "fluke".

yeah it seems that the disk is working normaly ....the graph is excelent... you have something acessing your hard drive while is in testing like pclab says

FYI, although I no longer support Windows (and Windows support is off-topic for this forum), older Windows versions (e.g. Win2k) certainly didn't like having a drive and its clone (i.e. same Windows drive signature), in the same system at the same time. I don't know if Win7 (as you are running) handles that situation better, but I just caution you that your config of having both drives (original & clone) in the same system at the same time, and booting from one of them, concerns me. If you've already checked with a reputable source and this is a supported configuration on Win7, then please ignore this worry