If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
|
Thread Tools | Rate Thread | Display Modes |
#16
|
|||
|
|||
O/T: Win7 m.2 2280 clone to hdd.
J. P. Gilliver (John) wrote:
In message , Paul writes: J. P. Gilliver (John) wrote: Is there any way to tell whether a drive is a (firmware-based) one of this type, so as to be able to avoid them? Using "shingled" or "SMR" is for public relations purposes a poisoned descriptor. You're not likely to find an admission of which ones are shingled. [] )-: Is my "HGST HTS541010B7E610 (1000G)" (really 931 GiB of course), bought over the counter a few months ago, of that type? Can I tell from any part of that number? Let's try a little experiment. I think WDC owns HGST now. HGST used to be IBM Research. This is your drive, with the 19 and 21 dbA acoustic properties. https://www.hgst.com/sites/default/f..._datasheet.pdf Compare to the middle column here. https://www.wdc.com/content/dam/wdc/...879-771437.pdf I think it's the same drive. Notice that only the middle drive has a 128MB cache. The others have 8MB and 16MB cache (likely older controller boards). Only the slimmest drive got the big cache. Makes you wonder... Hmmm. ******* The exercise requires a *lot* of supposition. https://forums.anandtech.com/threads...tters.2525313/ Even if the idiots told us how many platters, that would help. They don't even give areal density with regularity. There's just not enough data to work it out - my data comparison method is no damn good, unless you can trace down the release date on each drive on Page 2. A drive design could have a really small cache, if it was released ten years ago. More than a little weird, as modern memory chips are huge, and you'd probably have to pay a premium to get some crusty old 8MB chip. One of the consequences of buying a small chip like that, is the bandwidth might not be that high either. Paul |
Ads |
#17
|
|||
|
|||
O/T: Win7 m.2 2280 clone to hdd. Now shingled versus ordinary drives.
In message , Paul
writes: J. P. Gilliver (John) wrote: In message , Paul writes: J. P. Gilliver (John) wrote: Is there any way to tell whether a drive is a (firmware-based) one of this type, so as to be able to avoid them? [] You're not likely to find an admission of which ones are shingled. [] )-: Is my "HGST HTS541010B7E610 (1000G)" (really 931 GiB of course), bought over the counter a few months ago, of that type? Can I tell from any part of that number? Let's try a little experiment. I think WDC owns HGST now. HGST used to be IBM Research. This is your drive, with the 19 and 21 dbA acoustic properties. https://www.hgst.com/sites/default/f..._datasheet.pdf Thanks. Compare to the middle column here. https://www.wdc.com/content/dam/wdc/...ssets/eng/spec _data_sheet/2879-771437.pdf I think it's the same drive. (Well, second-from right column now - the middle one's a 1.5T drive.) Notice that only the middle drive has a 128MB cache. The others have 8MB and 16MB cache (likely older [] The exercise requires a *lot* of supposition. https://forums.anandtech.com/threads...ives-have-smr- platters.2525313/ That does say the bigger cache _might_ suggest SMR (as SMRs _need_ bigger caches). [] design could have a really small cache, if it was released ten years ago. More than a little weird, as modern memory The manufacture (as opposed to release) date on my drive is late last year, if that's relevant. [] So, in short, it's hard to tell. So far, the drive seems to be working fine (running almost continuously, though not heavy use - Windows 7, and I don't do a _lot_ of number-crunching). Is there any failure symptom that shingled drives would exhibit that is different to Pxx (I forget the letters) drives, that the ordinary user could detect as being different? Or just general increased failure probability? -- J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf Lucy Worsley takes tea in Jane Austen's Regency Bath. - TV "Choices" listing, RT 2017-5-27 |
#18
|
|||
|
|||
O/T: Win7 m.2 2280 clone to hdd. Now shingled versus ordinarydrives.
J. P. Gilliver (John) wrote:
Is there any failure symptom that shingled drives would exhibit that is different to Pxx (I forget the letters) drives, that the ordinary user could detect as being different? Or just general increased failure probability? They're made from the same materials as a PMR drive. Just the patterning on the platter is different, so the servowriter job at the factory is harder. In terms of flying height, head shape, I think they rely on the low fringing of the PMR design, to make the SMR pattern work. It wouldn't be a failure symptom as such - it would be the "flaky feeling" on writes that would annoy. Normally, we get a flaky feeling from reallocated sectors of disk. Added to that, would be the need to write seven track chunks for the "SMR bonus". With an SMR design, you might feel less in control of what's going on. "Is my drive bad or is it having one of those days?". It makes it harder to answer the "why is my drive slow" question. I'm sure it'll have a SMART reallocated parameter like other drives. Just a guess. Paul |
|
Thread Tools | |
Display Modes | Rate This Thread |
|
|