Single platter 500GB HDD

So OK already I ordered a 7mm thick HGST Travelstar 2.5-Inch 500GB 7200RPM SATA III 32MB Cache for $78 from Amazon.

FWIW given 7mm thickness, I'm assuming it's a single 2.5" platter drive.

Given single platter, the transfer speeds should be slightly improved over 500GB dual platter drives.

 

Only if you don't partition and use it full capacity.

If a single platter vs. a dual platter drive is partitioned (to say 100GB 'my' standard) then the dual platter will be faster in transfer speeds and in responsiveness.

This is because the heads are actually moving half the distance of the single platter drive (and of course assuming everything else stays the same - including the data density).

 

Interesting thought process, but unlikely to apply to me.

I'll be using the drive strictly to backup my 512GB M4 SSD. The M4 SSD is in the main bay, and the new HGST Travelstar 2.5-Inch 500GB 7200RPM drive is in the DVD bay.

Both bays are SATA III capable, and both the M4 SSD and the new HGST HDD are SATA III capable.

One open question is if a single platter 2.5" 7200 HDD doing constant writes can now exceed SATA II speed limits?

But that is strictly a peanuts issue to me. Meaning that it is at most mildly important to me.

The acid test of my setup will come when I pop out the SSD from the computer, and pop in the cloned backup HDD and try booting it with the HDD without changing any of the BIOS stuff.

If that works, then I've got industrial strength backup with a SSD setup.

 

That's not industrial strength - there are many things that can go wrong with your method (#1 being that the clone is not a recent one).

What I would be aiming for is this:

O/S:
Win8x64 PRO

SSD:
100GB C:\ O/S and Program drive - more than enough for all programs you want to install (forget steam - that is 'data' to me).

256GB D:\ - Data drive - expand as your needs indicate (up to the full capacity of the SSD) - this is where 'steam' games go to.


HDD:
100GB C:\ O/S and Program drive - yeah Dual Boot with a fully installed copy of necessary Windows and Programs.

256GB D:\ Data drive - this is where the SSD's Data drive gets backed up to (SyncToy would be my preference).

109GB E:\ File History drive - this is where (either) the SSD's Win8 install does automatic (aka timemachine-style) backup of your Libraries.


Sure - the above could have some of the same problems as your method (copying of viruses, for example) but it avoids backing up system errors in setup and you can dual boot at any time to verify one O/S' operation with the other - without removing/switching your drives.


And, while no mobile HDD can exceed (even remotely) SATAII limits - the same drive will still perform faster (more responsive) in a SATA III port.

 

Very sorry to snip to this, but man there is a serious communication failure here!

And re writing 100GB of backup data on a SATA III HDD, I seriously doubt that it'll be any faster than a similar SATA II HDD.

Being "more responsive" means stuff that is pulled randomly from the large cache that the HDD has.

Which does not apply to a 100GB backup situation.

 

Hey, you don't have to believe me - try, use, experience different setups and platforms and 'grok' this for yourself.


I am not going to find the post - but I (fairly recently) posted how a SATA I port took 10x longer than a SATA II port on the same system... (I don't expect SATA2 to SATA3 to be much different).

 

Wow this is a rapid fire exchange between us.

I'll put a stopwatch on the new single platter drive when it shows up.

I'll put a stopwatch on a two platter WD Black.

I expect the single platter HDD to have a slightly lower backup time.

I'll post the honest result here, one way or another.

What's gotten lost in this exchange is the main issue (to me) of if I can pull out the SSD drive and drop in the cloned backup HDD drive (of a resident SSD drive) and have the laptop boot OK from the backup HDD.

Having a slightly faster clone backup (due to the single platter drive having higher density) is a small plus, but only if the stopwatch proves it.

 

That test won't prove anything one way or another: you're comparing two different drives not single vs. dual platter.

As I said - if leaving the identical (manufacturer/firmware/model/generation/etc.) drives un-partitioned (except one single platter and one dual platter) then there won't be a measurable or noticeable difference.

But using HDD's un-partitioned is like using a two wheeler with training wheels.

 

Single platter drives always have the edge in sequential transfer speed due to higher areal density vs their multiplatter counterpart of the same capacity and RPM. This is due to the head picking up more data per revolution on the single platter model, this performance is also maintained reasonably well as the drive fills up. Single platter drives have higher reliability simply due to the smaller number of mechanical parts and heads. Single platter drives also use less energy and output less heat + noise due to the lighter load the motor has to spin.
From what I've seen, high RPM drives are only good as system drives due to their lower access times hence faster random performance but their sequential (i.e. backup) speeds are about on par with a high density low RPM drive due to the usage of lower areal density platters. The serious drawback of high RPM drives are lower reliability and higher heat output (which may be important depending on your enclosure) therefore, I don't think it is as suited for a backup.

What tiller spoke of is short-stroking the drive to limit the range the head has to move, in theory, you get a higher minimum sequential speed too by not letting data go to the inner platter tracks. The multiplatter (due to more heads in tandem) and high RPM drives gain an edge in random performance here.
In your situation, the 500gb drive is more than sufficient for a system image backup, I have a WD 1tb caviar blue in an external 6gb/s enclosure (2x500gb platters) which stores the system images for my SSD.
If you want a direct clone, then you gotta be careful of alignment, I had issues migrating Win 7 from an HDD to the Intel 320 drive because the MBR got totally wrecked by Acronis. I did a disk repair and startup repair and all was good.

 

Ignore tiller. What he/she says about storage is irrelevant to 95% of the computing populace at large.

 

And to assume "everything" else stays the same is your logical flaw. Such a flaw as that, well, silly.

It really takes almost no effort to go on the Hiatachi site (or whatever they call it now) and find the areal density of their single vs dual platter. And that will confuse you even more. Why? Because you have 3 drives of different sizes and the same Areal density and the same stats same line. Rotational speed is also in the mix. They control the movement of the heads on the different capacities.

Tiller on your two platter less movement? I get what you are saying but wrong. In a confined space one step forward and one step back is still two steps. Two steps forward is two steps also. I might give you random access but not even sure on that.

To OP yes you have the single platter. Yes it is their high performance model. So congrats and good you have what you want.

I am done with HDD's for performance. In a few years I think everyone will be also. SSD's are the future for performance. HDD's might always have a role in mass storage but who knows, I don't.

 

Good comments.

And yes, Tiller also came in with good comments. Certainly got me thinking.

I never had any doubt on that.

Is that WD 1TB Blue a 3.5" HDD?

FWIW my long experience with 3.5" HDD (mostly 7200 rpm this past few years BTW) has always been that higher platter density = higher performance. But they've all been multi platter drives.

Alignment was an issue with cloning with the Acronis True Image Home 2010. But they fixed that with their 2011 release, which is what I still mostly use, current Clonzilla being the other. Another thing was that Acronis seemed to offer rolling updates (to the CDs in the boxes) as the model year progressed, but their support people would never admit that, which I hated.

IMO Acronis support is like dealing with the lowest weasels. But it's possible I might try their current package if I come up needing something specific to work with a SSD setup. One step at a time.

Cloning properly, from a HDD to a backup HDD, has the great advantage that the backup HDD can be dropped straight into the machine and presto you've got your Windows kluge back and working. That's saved my butt more than 3x this past 15 years. Given how much time goes into setting up a working computer system is why cloning is always my 1st choice. Imaging is faster and more convenient, but it doesn't really cover worst case scenarios.

 

I can tell you have no experience with this.

If you don't keep everything else the same you are comparing apples and oranges.

Spec's don't indicate what performance you'll get between models - only what the manufacturers aimed at ('scores' do not equal real world results).


On a dual platter system the outer (fastest) tracks are read/written to first. The heads don't need to move - they simply read or don't read on a given platter/side per rotation.


You may think my logic is flawed (and others can continue to think that my storage sub-system experience is irrelevant to most) but that just shows your inexperience with this topic, not mine.

Also, areal density is not higher on 'only' single platter drives - take the same platters with the same density, controller, firmware and cache and dual platter drives partitioned properly outperform single platter solutions when the highest sustained minimum transfer rates (and responsiveness) is required.


The OP asked a specific question and I answered it in full.



To put my original answer to the original question another way:

If a 1TB HDD (otherwise identical to the 500GB HDD except being dual platter) was used and partitioned to 500GB - it would be superior to the single 500GB HDD because the minimum transfer rate would be much higher and the responsiveness of the system would be obviously improved too (because the heads are now traveling half the distance across the platters).

This increase in minimum transfer rate and increased responsiveness continues until somewhere around 10-20% (depending on the specific HDD) of the capacity of the drive is used as the first partition (for O/S and Programs) - in other words - up to about 100GB capacity for 500GB HDD's (~20%) or the same 100GB capacity for 750GB HDD's (~15%) or the same 100GB capacity for 1TB HDD's (~10%). What changes in each scenario above is that the heads need to cover less and less area for the same sized partition and in this way ensure higher performance.


With regards to the additional information the OP added in later posts: the above is still valid - just that there is no reason to apply it to his/her situation. But that is not what was originally asked.


You may want to read some of the links here if you really want to know more:

See:
http://forum.notebookreview.com/har...ades/441674-so-many-tests-so-little-time.html



So while you think that keeping other variables constant is 'silly' when trying to determine if one of two options offers a significant performance gain - keep in mind that it is the basis of all scientific progress. Without which we couldn't have intelligent conversations or repeatable results without.

 

The drive is 2.5inch, essentially identical to yours except I can fill it up to twice your capacity with the same performance dropoff curve. I hated that the high RPM drives (excluding Raptors) tended to have lower density platters so the sequential performance advantage was not as pronounced. I still have a 1tb WD RE2 chugging along, I can only imagine how many platters that beast has. That being said, I'm eagerly waiting for 1tb single platter 2.5 inch drives.

I'm never using Acronis again lol, all my systems are now on SSDs so I hope to god I never have to do another HDD-SSD migration. I use mostly Windows image backups, they are really streamlined with Windows 8.

 

So OK, I got my new HGST Z7K500-500 7mm 500GB 7200 HDD yesterday, and this morning did two clone backups.

Results were just barely better than before, so at best a very small improvement for sequential xfer of data. However it's not clear where the bottleneck is.

Laptop: Lenovo 15.6" ThinkPad T530 i5-3320, from main bay HDD (SATA_III capable) to ultrabay HDD (SATA_III capable)
SW for running the clone: Acronis True Image Home 2011, booted from external CD drive connected by USB 2.0
Source: recent WD Scorpio Black 500GB 7200 9.5mm 2 platter, SATA_II capable

1. Target: recent HGST Z7K500-500 500GB 7200 7mm single platter, SATA_III capable
2. Target: recent WD Scorpio Black 750GB 7200 9.5mm 2 platter, SATA_II capable

Data: 3 NTFS primary partitioins, totalling 78.3 GiB

Clone backup timings:
1. 14 minutes 35 seconds rounded up to next second
2. 14 minutes 38 seconds rounded up to next second

Useful 7mm 2.5" HDD review refs:

Hitachi Travelstar Z7K500 Review | StorageReview.com - Storage Reviews

Western Digital Scorpio Blue Review - 500GB 7mm (WD5000LPVT) | StorageReview.com - Storage Reviews

 

If my math is correct, that is just under 92 MB/s - I wouldn't have expected any better with a non-partitioned drive(s).

 

Keeping in mind that 1GiB ≈ 1.074GB, I'm just gonna use GB.

So 84.1GB/14.58min = 5.77 GB/min, which is pretty good for laptop 2.5" HDD.

Best I've gotten on a desktop PC w/3.5" 7200 HDD (18 months ago) is low 6+ GB/min

 

Your source hdd was proably slower than either targets. Thats probably the bottlneck. The only way to truly gauge the difference is to have an SSD source.

 

That's a very good point.

In about 2 weeks I plan to put a 512GB Crucial M4 2.5" SSD into the main bay (as boot drive), and do a fresh install of Win7. I've already got the SSD.

So I'll update clone timings again, when I get the SSD as the new faster boot/source drive in there.

 

I just did my 1st clone of my main/boot SSD (nominal 512GB M4 2.5" SSD) to a temp 500GB HDD (same HGST HDD that I used a month ago, see above refs) in the DVD bay, on my new T530 laptop with a single NTFS partition (fresh Win7 installed by me); then pulled out the boot SSD and dropped in the HDD and it booted fine.

I used Acronis True Image 2013 CD (clone > manual > "as is"), which I booted the laptop with (NOT installed into Windows); I also have a CD of True Image 2011 (from late in the Acronis 2011 model year) and am fairly certain that it would also work with that or 2012 version.

The M4 SSD is SATA III and the temp HDD is a 500GB single platter (7mm thick) HGST 7200rpm which is also SATA III, and both bays work at SATA III speed.

It took 8.6 minutes to do the actual clone backup (not counting time spent moving drives around).

So 58.71GB of actual data on the partition / 8.6 minutes = 6.83GB/minute, which is even better than the best clone xfer rate on my fastest desktop.

 

Thats more like it. I'm getting a Samsung 840 500gb for external backups, curious how fast I can get it between that and my 256gb Samsung 840 pro

Out of curiosity, could you do a HDtune bench on the HGST drive for me? I want to see how it compares to my WD 1tb Caviar blue 2.5inch drive.

 

Seems like a good pick to me, given the low current price. And Samsung's reputation as a SSD maker is in the ballpark of Intel and Crucial.

Is this via Newegg's current $280 deal?

HDTune is a several year old benchmark (version 2.55). Why do you value that one???

Anyhow I have it and will run it and post a pic of the result. Sometime later tomorrow.

Given that my T530 is likely different than what you're running, it may be like comparing apples to oranges.

BTW I'll also run a 2nd identical clone backup, but with using a 2.5" WD Scorpio Black as the target drive (7200rpm but "only" SATA II).

The target HGST HDD was just barely faster than a target WD Scorpio Black, when the source was a 2nd WD Scorpio Black, and I kinda suspect that the same thing will happen again. Meaning that the reason that this new clone backup went somewhat faster was having a SATA III SSD as source drive (i.e. 512GB M4 SSD).

 

I find HDtune tends to tease out the true performance of a HDD even when it already has data. Plus, it is fairly realistic as it also depicts performance as a function of capacity used. I don't like it for SSDs though as it burns through P/E cycles like anything.
This is the WD 1tb 2.5 inch Caviar Blue drive over USB3.0 with my desktop (which has a 256gb Plextor m5s over SATA3 with an I5-2500k @4.4ghz) so hopefully no random bottlenecks

Higher peak with dual high density platters but drop-off is massive with capacity

Plus, also my ancient Storage WD RE2 1tb 7200RPM (this monstrosity has a ton of platters so it sustains performance as it gets filled up quite well)


Funny how technology improves huh?

 

OK on HDtune. Nice pics (that I snipped). Amusing last line.

w/ regard to HDtune, in the upper right corner is an "options" icon. Do you leave it at default (somewhat mid way), or set it to max accuracy???

Also same question for block size, which has default (on my desktop PC) set to 64KB.

 

Standard, settings, 64kb block size. I tried the high accuracy but didnt seem to make much difference

 

How can you have the nerve to say something like this when you are the poster boy for unscientific conclusions? You exaggerate everything, ignore criticism, ignore evidence that contradicts you, and claim you base your conclusions on some most definitely unscientific data you gathered yourself.

 

What evidence? And where do I contradict myself?

Real world (actual) results always trumps 'theory' - no matter how sound it may seem at first blush.

Choosing to ignore uncalled for criticism is something bad now? Perhaps even in posts like your own?

If you can leave personal attacks in your head and come here with a real topic to discuss, then I'll gladly engage in that conversation.


Please find something more useful to do with your posts than attack me personally all the time. Thanks.

 

Results using my T530 laptop

source= 512GB M4 SSD, SATA III, in main bay which is SATA III capable; DVD bay is also SATA III
target1= HGST 500GB 1-platter HDD 7200rpm AF (Advanced Format) SATA III: 8.6 minutes to do clone
target2= WD Scorpio Black 500GB 2-platter HDD 7200rpm not AF SATA II: 9.0 minutes to do clone
target3= WD Scorpio Black 750GB 2-platter HDD 7200rpm not AF SATA II: 8.7 minutes to do clone

So for sequential writing, the HGST wins, and by a slightly greater margin due to the faster source than my OP of this thread.

5 HDTune pics:

HGST 500

WD Scorpio Black 500

WD Scorpio Black 750

M4 on fresh boot

M4 on 3rd try without rebooting