I had not realized how close 3.5" drives had come to the noise level of 2.5" drives. Its almost to the point that traditional 2.5" spinning drives are quickly becomming a niche market.

You've got 3.5" drives with a better price/GB that are just as quiet for mass storage, and you've got SSD's with better performance/acoustics/battery consumption for your notebook/desktop OS drives.

The only thing traditional 2.5" drives have right now against SSDs is price, but this wont last much longer.

That's not quite the whole picture, imo. The number of cheap small computing devices is not going down, it's accelerating. The Dell Studio Hybrid is a prime example. A 3.5" drive has no place in such devices -- too big, too hot, too much power -- and at least for now, if you want any kind of capacity, SSDs are out of the picture. The 2.5" drive is perfect for such devices. The main question is how longit will take SSDs to come down enough in price to make them preferred in the mainstream.

Yes 2.5" drives will probably go the way of the Dodo sooner rather than later, but they do still have a fairly big market open to them.
Many people dont want a full size tower case on their desk or under their TV. Cases are getting smaller and more stylish, and in these cases a 2.5" makes perfect sense.
Ironically the laptop drive will probably disappear from laptops sooner than it does from PC's.

You mention areal density in the article comparing 394Gbits/in2 vs a 7200.11 of unknown density (the 7200.11 drives have much lower areal density, even 7200.12 drives have lower density). I spent some time last week comparing areal density between 2.5 and 3.5 drives just for my own curiosity so I can share those results with you.

2.5" Seagate mobile drives
394 Gbits/inch2 for 7200.4/5400.6 (250GB platters)
253 Gbits/inch2 for 7200.3/5400.5 (160GB platters)
204 Gbits/inch2 for ......./5400.4 (125GB Platters)
169 Gbits/inch2 for 7200.2 200GB model only (100GB platters)
131 Gbits/inch2 for 7200.2/5400.3 160GB and below (80GB platters)
??? Gbits/inch2 for 7200.2 50GB platters

2.5" WD mobile drives
??? Gb/in2 for 250GB platters (WD5000BEVT, etc)
250 Gb/in2 for 160GB platters
200 Gb/in2 for 125GB platters

3.5" Seagate desktop drives
329 Gbits/inch2 for 7200.12 (500GB Platters)
277 Gbits/inch2 for 7200.11 (375GB Platters)
228 Gbits/inch2 for 7200.11 (320/334GB Platters)
180 Gbits/inch2 for 7200.11 (250GB Platters)
109 Gbits/inch2 for 7200.10 (187GB Platters)
109 Gbits/inch2 for 7200.10 (166GB Platters)

3.5" WD desktop drives
520 Gb/in2 for 640GB platters (unknown 2011 series)
400 Gb/in2 for 500GB platters (WD20EADS, etc)
250 Gb/in2 for 320/334GB platters (WD6400AAKS, WD6401AALS, etc)
??? Gb/in2 for 250GB platters

It took a lot of Google searches to find some of these numbers. Most are from Seagate or WD themselves but they don't make it easy to find in some cases. Often Seagate will put the number in a technical PDF and a press release. WD usually only does it in a press release. Occasionally a retailer or review will have a number where the press release doesn't. Presumably the PR section of the company web site doesn't bother to archive every release or I'm just not thorough enough in my searches.

Because it was easier to find data on the seagate drives I was a little more thorough on that side. I gave up easier on finding data on WD 2.5" mobile drives.

It's probably worth stating that the manufacturers are probably rounding or fudging this numbers to some extent. Notice that the 500GB 3.5" platter from Seagate is stated as 329 Gbits/inch2 but the same type of platter from WD is stated as 400 Gb/in2. That is quite a range for what should be the same density platters. I'm guessing Seagate gives more accurate numbers and WD rounds up but there are other possibilities.

Maybe WD takes a higher density platter and leaves more spare sectors to improve reliability. Its very hard to say without having access to insider data.

From my own experience with HDDs, I can tell you that if hard-mounted, no 3,5" drive is even a close match for any 2,5" drive out there; however, when suspended, things change a lot and a quiet 3,5" drive may not be particulary disturbing. I have taken the extreme and suspended an antique 160GB Momentus 5400.3 in my desktop - I can tell you, it's inaudible from 1 m - but so would be a Caviar Green, I guess, with much more performance and capacity to offer.

As for the review - it comes at the right time for me, my music collection has been growing up steadily in size recently and it simply doesn't fit on my "old" 100GB Hitachi 7K200... Using an external drive proved to be tedious, so a 500GB upgrade was circulating in my mind for some time.

I guess I'd have chosen a Hitachi again, my current drive is amazingly quiet for a 7200 RPM drive, much quieter than most 2-platter 5400 RPM drives I've heard (Seagate, Toshiba, Fujitsu, etc.), and certainly much quieter than those nasty Momentus 7200.3 drives... Given the 12,5 mm height of the Hitachi 5K500, I may try a WD this time, they've got such nice color labels

what's interesting is how these high density 5400rpm 2.5" drives compare in performance and noise with 5400rpm 3.5" drives - performance is likely to be the same, noise should be similar (extrapolating down a bit from the WD6400AAKS to the GP WD6400AACS), but vibration should be better at an increased cost. So depending on available mounting options, that cost may be worth the improvements or it may not.

I'm a bit wistful about this review anyway - my old Dell has a 40GB PATA drive and larger capacity replacements are both thin on the ground and very expensive

Price doesn't seem bad to me, but availability is more the issue, esp as you go higher. 160gb for PATA 2.5" i easy to find. You might find the odd 200gb or 250gb. Here's a WD 250gb at newegg -- $85 seems a decent price to me: http://www.newegg.com/Product/Product.a ... 6822136159 Both Seagate and WD 160gb models are at $65.

thanks Mike, but sadly the situation's a bit different (as with most PC stuff) in Australia - 2.5" PATA drives command a ~25% price premium over SATA and the lowest price shops have stopped selling them completely. On eBay, 60-80GB drives go for ~AU$60-70 (about US$50?) and higher capacity drives are ~AU$110. Not a whinge, it's just the way it is

mattthemuppet --

I have to admit that after a couple months living with the Intel 80GB SSD in an Asus Eee PC (yeah total overkill, but there it was, looking for some kind of use), I'm no longer much interested in a disc drive for my laptops. I just picked up a used Lenovo x300 for a good price -- it's fitted with a 64gb Samsung 1.8" SSD. Still not the quickest PC, the C2D in there only runs at 1.2ghz, but nice how quickly it boots & apps open. Total absence of HDD noise means when the fan comes on, you really notice.

Great test!

I have my own experiences to share about this WD drive.

Now i've been following this great site for years trying to create a silent bedroom computer, and with this HD i'm finally satisfied. I have the WD 500GB mounted with Novibes 2.5" in a Silverstone TJ08. And again i can't hear the thing when i'm a sleep. It sits maybe 1.5meters from me. I also use Noctua 12cm fans, Silverstone STN03 fanless PSU and fanless Scythe Ninja mini in it. 4850e cpu.

I used to have a 3.5" Samsung mounted with Novibes in the same case. The problem was vibration and heat. Since i also wanted a small case like the TJ08 and not some big monster with much better vibration-dampening qualities, this hd was a winner and makes a difference. The heat is also very low. I can't notice the performance difference in every day use.

So thanks SPCR

I imagine that's where the mobile market is going - SSDs have so many advantages over HDDs that it's really just a case of waiting for the price to come down before they become more widespread. I can see 2.5" HDDs hanging on in either desktop replacements, which need the extra storage, or bargain basement laptops.

Hmm, 500 gb 3.5" platters can't be both 329 and 400 gbit/inch2.

BTW, new 1 tb drives can do 100 - 120 mbyte/s.

http://www.storagereview.com/php/benchm ... 9&devCnt=3

Mike,

Maybe you could find a way to start recording temperatures at idle and after HDTach runs? Might be good to know the temperature limits of some of these drives. Particularly the notebook drives, as their temps will get much higher when installed in a notebook chassis.

The temp rise would be extremely small, as HDTach last only a couple mins. I've never even noticed them getting warm. The thing to remember is that almost all notebook drives produce virtually the same amount of heat, the technology is that optimized & widespread: 1w or less at idle, maybe 2.5w max. The actual temperature reading will depend on so many factors -- location of sensor, ambient temp, applications, and of course installation details. I don't think recording temps would be very useful, just more white noise to sift through. The power consumption numbers are far more definitive.

The postscript says:

[quote]
SSDs' usability continues to improve, with performance besting the fastest of current disc drives. At time of writing, an excellent 128 GB Samsung based SSD is selling "on the street" for little over $300
[/quote]

Before you rush into buying a SSD as device besting HDDs in performance, read this:

http://www.anandtech.com/storage/showdoc.aspx?i=3531

It seems there's only couple of SSDs on market that have a good controller. And the rest will create annoying 1-2s full stops to the system once they have been used for some time.
[/quote]

You misread. Where does Anand say this? I came across not a single mention of such a thing. His verdict very different from your interpretation.[quote]
The Verdict

There’s no skirting the issue: even the best SSDs lose performance the more you use them. Eventually their performance should level off but what matters the most is how their performance degrades.

In using the X25-M I’d say that the performance drop was noticeable but not a deal breaker - and the data tends to agree with me. With average write latencies still well under 1ms, the drive maintained its most important performance characteristic - the ability to perform random accesses much faster than a conventional hard drive.

Keep in mind that with the cost per GB being as high as it is, these SSDs aren’t going to be used for large file storage in a desktop or notebook. You’re far more likely to use one as your boot/applications drive. As such, what matter the most aren’t peak transfer rates but rather fast access times. On a well designed drive with a good controller, peak transfer rates may fall over time, but latency remains good.

You end up with a drive that still manages to be much faster than the fastest 3.5â€

From the article, page 2:



http://www.anandtech.com/storage/showdo ... i=3531&p=2

So the key point he is making there, and the point of the whole article, is that you want to stay out of the JMicron controllers when going for SSD.

And that was my point, too: Be careful when picking you a SSD device. No sane person would re-copy all his data like once per month..

Like Anand says:



Ie. without having an OS that would optimize the writes for SSD, there's a huge different between vendors about how their drives perform in practise.

Sorry to be rude, but no, you skimmed over the article too quickly. Anand specifically, repeatedly mentioned SSD drives with JMicron controllers (which is currently the most commonly sold kind), and problems with them. The abysmal performance of those SSDs in random writes, compared even to 5400rpm hard disks, is clear in the charts (see this page. Note that he only actually recommended the Vertex (and with a reliability caveat), and the Intel drives. I bet he'll give the new Samsung ones a spin, since they promise improvements as well.

Of course Anand was talking about stutter, that was the core of the whole article, give me credit for reading! It's this comment that I said is wrong:


If they stutter, they stutter from the start. Obviously, you want to choose a drive that doesn't stutter, but the wear/slowdown over time is definitely solvable.

Negative.

To quote Anand:


And:



The SSDs are fast until the point all the areas in the SSD are written once. Because this is something the controller aims to hide from you, you cannot know when exactly it happens, but the more you write, the sooner it happens.

And that's the message of the said article: the wear/slowdown is solvable, by use of good SSD controller circuit. The another message is that some are good, some are bad. So when shopping for SSD, be sure to read the said article and the differences between different controllers described there.

I didn't ask anyone to take my word for it, I asked you to read the said article before shopping.

zds --

We're both just citing Anand here. I'm not sure why you don't see what I see:

If they don't stutter at the beginning, then they don't stutter, period. The ones that stutter, well, why would you buy one in the first place?

Even if they slow down over time, [quote]You end up with a drive that still manages to be much faster than the fastest 3.5â€

That's the problem. No SSDs stutter right from the start. Good drives become a little slower when they have no more fresh free space for writes. Bad drives (the JMicron controlled ones) start to stutter at that point. That hidden "feature" of cheap SSDs is the problem Anand wrote an earlier article about, and he repeatedly stated it here. You missed the distinction between cheap/older SSDs, and the more advanced ones.

Sorry, that's not what I read. You'll have to convince me it's what Anand wrote, but what I recall of his first article was that the SuperTalent, Silicon Power and OCZ Core SSDs with JMicron controllers stuttered in XP/Vista right from the start. It took him a number of tries before he devised a test that brought out the worst, but the stuttering was there right from the start, and it's what he didn't like about the earlier SSDs he was trying.

With regard to the performance decline over time (in simulated heavy usage), in PCMark Vantage HDD Tests, Anand found the worst drop of 36% on an OCZ Summit, but even then, the performance remained ~3x faster than the fastest desktop drive, the WD Velociraptor.

How could I possibly miss that -- and why would you say such a thing?

So back on topic, i upgraded the hard drive in my 2009 13" macbook pro from the stock fujitsu 160gb to the 500gb Scorpio Blue. It is fairly quiet, ill give it that. but it does seem a tad bit louder than the stock fujitsu hard drive. Are apple's hard drives quieter? or is it just because of the fact that its a 160gb hard drive, that its able to be quieter?

I have also realized that when I put pressure on the case (where the hard drive is) the sound goes away. so perhaps its just vibrations, but it doesnt really sound like it. I guess it sounds like what people have described as "putting ur ear next to a sea shell. Anybody have any ideas on how to possible fix this? add some foam between the drive and the case perhaps?

Has anyone tried the single-platter versions of these drives yet, and is there any improvement in power or vibration (especially in the WD)? I'm a little leery of Seagate at the moment due to some bad experiences with the Momentus 7200.4 (supposedly same platters, but 7200 RPM), so I'm kind of leaning towards the WD. What is the right model number to look for to ensure I'm getting the latest gen 250GB drive? Would it be the WD2500BEVT?