17 Serial ATA Hard Drives Compared 133
TheRaindog writes "The Tech Report has an in-depth look at Maxtor's DiamondMax 11 hard drive that provides some interesting insight on how Seagate's recent acquisition can improve deficiencies in its own drives. More valuable, however, is the fact that the review offers a detailed comparison of 17 different Serial ATA drives from Hitachi, Maxtor, Samsung, Seagate, and Western Digital. Performance is compared across a wide range of typical desktop, multitasking, and multi-user loads, and noise levels and power consumption tests also provide interesting results. Definitely worth a look for anyone in the market for a new hard drive."
Loud noises! (Score:4, Informative)
Re:Loud noises! (Score:5, Interesting)
The measurements were made one inch away from the drives, so you could expect the levels to be pretty high. Unfortunately, their measurement methodology also means that the noise levels are useless for anything but comparison purposes.
It would be nice to know whether the levels were A-weighted or linear. Also, with the meter they were using, differences of less than 2 dB aren't meaningful.
Re: (Score:2)
not really (Score:5, Informative)
Re: (Score:3, Informative)
Re: (Score:2)
If you like this sort of thing here's another wiki about equal loudness contours (basically how our ears react to different freqs and levels - it's a whole more complicated stor
Re: (Score:2)
Well, the base of the log _is_ the factor between the 10-dB increments. It's a lot easier than having a scale where the each increment of 10 means 2.7182818284590451.. times more power. E-based logarithms are more natural for scientists, but the dB scale is supposed to be something you can explain to laypeople.
As you can see from the Wikipedia article, this was originally called the 'bel' scale in honour of Alexan
Data (Score:2)
I was not too happy to learn of the merger~boyout, and while the article hints at optimism, I definitely get the sense of a rolling of eyes out there. Competition is what spurs creativity and success. We shall see.
Missing statistic... (Score:5, Interesting)
I'll gladly sacrifice a few percentage points of performance if it means increased reliability, especially when we're talking HD's. I already don't trust the things farther than I can throw 'em (thus the RAID-5).
Re: (Score:1)
Can't use those reviews as a real indicator (Score:5, Insightful)
Re: (Score:2)
Re: (Score:3, Insightful)
In the case of the RE2 and newegg, I used the reviews as a guide, but when I searched the "broader" internet for people having problems with that drive, the "big" problem was just not to be found. As with anything and especially the internet, bu
quiet is nice, fast is better - but really... (Score:2)
Re: (Score:2, Insightful)
In fact, Newegg is absolutely littered with thousands upon thousands of reviews from people who return to rave about their new toys... a bit to the extreme, actually.
Go see how many HUNDREDS have come back to the site to post about their shiny new floppy drives, for crap's sake.
Back to the point -- 15/27 is still bad (esp. after reading the comments about failure) compared to the competition.
Not always the drives' fault (Score:2)
"...I wrote the previous review about 1/3 drives failing after two weeks of use. Turns out it was due to a low-quality power splitter that caused the drive to go up and down enough times for the RAID controller to mark it as FAILED. I used the extended test (about 2 hours) with WD's too
Re: (Score:2, Insightful)
Re:Missing statistic... (Score:4, Interesting)
The SATA standard is entirely suitable for enterprise use and has a growing collection of enterprise class drives.
Reliability is a "big issue" for almost everyone. Reliability is more important than raw performance for most of the desktop market, indeed most of the hard drive market. If SATA drives were inherently unreliable, they would be unsuitable for any market.
The only reason reliability isn't the main benchmark for a drive is that it's so hard to measure or predict. If measuring reliability were as easy as measuring noise levels, it would be the most important buying criteria bar none. Particularly since other measurable criteria do not vary all that much between modern drives at a given price point and capacity level.
Is SCSI RAID faster than SATA RAID ? (Score:3, Interesting)
My Hypotheses is simple:
1. What really matters for a RAID implementation is Reliability, Size, cost and Speed. In that order.
2. SATA drives come close to SCSI drives in individual performance. Greater data densities help and lower spin
Re: (Score:3, Informative)
Re: (Score:2)
Same could be said for SATA. As you said, the standard isn't going to make a difference.
"The premium on the SCSI hard drives themselves is justfied."
Depends on your priorities. Drive manufacturers want the high margins on enterprise drives so they deliberately differentiate between SCSI and ATA. It's mostly a marketing ploy.
Re: (Score:2)
This is exactly right. I run about 50 TB of RAID drive sets under extremely heavy 24/7 load, and the SCSI drives and RAID controllers outperform the SATA drives and controllers by about 4x. This has been mystifying, because by the numbers the SATA drives and the SCSI drives should perform almost identically. My current theory is that the SCSI RA
Re: (Score:2)
Re: (Score:1)
Keeping drives cool makes them last longer. (Score:2)
Re: (Score:2)
It isn't SATA vs SCSI that determines performance, it's the class of drive that does (and enterprise drives are still SCSI).
SAS is NOT far superior to SATA. In fact, SAS uses SATA's physical interface (deliberately) and implements the SCSI protocol instead of the ATA one. They did that because the SATA interface was perfectly
Re: (Score:2)
2. That's an observation based on current product offerings. It is not inherent in the interfaces.
3. Assuming the drives are from the same family, the larger the drive the faster it performs. If you are comparing different drives then no general statement can be made.
By posing this question, are you asking others to do your research for you or are you asking them to do your lab work for you. It seems to me you should be doing thi
Re: (Score:2)
So, yes, I _AM_ asking others to do my lab work for me.
I am crafting a more detailed version of the parent post to send to people who do this kind of research for a living. I.e. ZD, Tom's hardware etc...
Re: (Score:1)
Spin rate is not a function of SATA or SCSI, as you know, so is moot. Most SCSI drives are still 10K RPM, so Raptor 10K SATA drives are of a similar performance, and considerably more expensive than regular SATA drives.
Latency dominates performance for most workloads other than pure strea
Re: (Score:2)
Re: (Score:2)
Re: (Score:2, Informative)
The difference is very stark, in terms of drive failures. We have a seriously disproportionate number of SATA drives fail, to the point where we simply aren't buying BL35p's anymore with SATA, they're just not worth the extra hassle from drive f
Re: (Score:2, Informative)
I work in a heterogenous hardware environment (with just hundreds of drives, not thousands
But mo
Re: (Score:2)
That said, the proper response is to stop buying the products that fail as you've done. That doesn't mean that SATA is not worth owning however.
Re: (Score:1)
Indeed, and SAS (Serial-Attached SCSI) drives differ mainly in rotational speed, # of devices on a bus, and warranty.
SATA definitely has a place in the datacenter (local disk, archive/bulk disk tier).
Re: (Score:1)
Dude, I said I'd sacrifice a few points of performance, not a few hundred bucks!
I'm also gonna have to agree with Sparohok here: "If it was a big issue, you'd be using SCSI"... that is so bogus.
For my home RAID fileserver, and many other applications, SATA is more than enough. Besides, I've seen my fair share of SCSI drives go Tango-Uniform on the job, too.
Re: (Score:1)
You want a reliability review? How exactly would you yourself go about doing a reliability review?
I'd have to get at least 4 each of these 17 drives, and test them for 6 months at least. Even then it's be far from a true sample, though it might offer some insights. True reliability testing is long, expensive work - perfomance testing can be done quick and cheap comparatively.
Also, just because 1 drive fails doesn't mean the drive series is unreliable. I personally like maxtor - I've used and sold abo
Re: (Score:2)
I've read some of those reviews, and it seems many folks are misunderstanding and/or misapplying WD's RAID Edition drives, presumably because they haven't read the datasheet [wdc.com]. Put simply, they are not intended to be used as single drives or in RAID0 configurations, and will be less reliable than regular 'desktop' drives (e.g. WD's 'Special Edition' models) in these roles. My understanding
Re: (Score:2)
We have over 70 of them spinning at work in various servers and desktops,
about 20 are older than 3 years. Models range from 40 to 300gig.
And here comes the kicker: In my 4 years on the job only *one* seagate failed.
In my little failed-drive box I count:
7x IBM
5x WD (50%)
4x Exelstor (100%)
4x Hitachi
1x Samsung (2,5")
1x Seagate (80gig)
Obviously this is purely anecdotical and we're not running an equal
number of drives of each brand (actually we had only 10 WD total...)
but nonetheless whenever someone
flash ram drives (Score:5, Interesting)
It would have been interesting had they done a comparison with one of the Asus Z62F solid state machines that uses flash ram as a hard drive.
Seconded. Mechanical drives are on the way out (Score:2)
Re: (Score:2)
Re: (Score:2)
Why is this modded interesting? Solid state RAM drives aren't even close to being in the same market segment as SATA drives.
I can summarize the comparison:
RAM Drives are the best by a huge margin in every metric except for size
Ram drives are:
zero sound
zero latency
uber-fast seek times
less heat output
etc etc etc
Re: (Score:2)
No? You'd think nobody would bother making SATA flash disks then...
http://www.m-systems.com/site/en-US/Products/IDESC SIFFD/IDESCSIFFD/Products_/SATA_Products/FFD_25_Se rial_ATA.htm [m-systems.com]
Re: (Score:2)
Although they are certainly interesting, they definitely aren't in the same market segment as consumer SATA drives. It took me a while to find someone that sold them:
http://www.wdlsystems.com/modperl/view_services.cg i?r=list_aisle.plate&aisle_id=778 [wdlsystems.com]
In small quantities (less than 50), price per unit ranges from $711 USD for a 1GB drive to $3,215 USD for a 32 GB drive. If you have to ask for the price of the 128 GB drive that is
Re: (Score:2)
Re: (Score:2)
He's not saying there isn't a market for them, he's saying that the people buying SATA flash disks has very different priorities to the people buying consumer SATA drives.
the dark side of flash ram drives (Score:2, Informative)
I'll take a reliable mechanical drive over a FLASH drive, thanks.
(PS. Thanks fo
Re: (Score:2)
I suppose that depends on how often your "important data" gets written. I can think of lots of applications where flash degradation shouldn't be a problem. (And yes, I can think of lots where it sure would be.) I wouldn't put /var on a flash, but maybe /usr or root. Actually, I would love to have a small flash drive to boot from, so I could completely devote my disks to LVM.
You'll also be blown awa
Re: (Score:2)
No, wait... what's that word that means the complete opposite of interesting?
Re: (Score:1)
Hazard Warning (Score:2, Funny)
SMART support? Pretty please? (Score:4, Interesting)
The Tech Report has an in-depth look at Maxtor's DiamondMax 11 hard drive that provides some interesting insight on how Seagate's recent acquisition can improve deficiencies in its own drives.
Like better SMART support on Seagate's side? I was stunned at how much more SMART capabilities Maxtor drives have compared to Seagates and others. It should almost be a crime to produce a drive that doesn't have a SMART compatible error log (which Maxtors have- you can query it and see when+what the last errors were, for starters.)
I've also been stunned at how BAD modern drives are; a client lost FOUR maxtor drives out of a 12 drive array in the space of 2-3 months, and we literally couldn't replace them fast enough (also, the idiots that he bought the system from TURNED OFF autoverify on the 3ware controller, didn't install the linux drivers, didn't bother updating the card's firmware, etc. That'd be PCs4everyone in Boston, FYI.) I had a Seagate PATA drive with barely a dozen hours on it that started clonking like crazy if you wrote data at high speed to it for too long (no, this was not the 7200.8, which had similar issues, relating to a motor driver circuit overheating. This was a 7200.9!) Seriously- the drive would completely stop writing data if you wrote to it continuously for about 40-50GB. The only thing that let me successfully complete the imaging was a borrowed fan directly cooling the drive.
I'm not too optimistic that Maxtor and Seagate will benefit each other in terms of technology the end user will care about; what is more likely is that Seagate will go enterprise, and Maxtor will go consumer, since that is what each brand is best known for.
Re: (Score:3, Interesting)
Re: (Score:2)
Re: (Score:1)
Re: (Score:1)
Re: (Score:2)
I had a similar situation but I really think it was because the drives were not rated to operate in an oven - which is what a badly designed case with 16 drives ended up as. That said it appears I've lost two Seagate SATA drives from a newer machine in the last two days, both halves of a mirror. I'm happy I fed it a new drive yesterday or I'd be looking at a long restore
What's most important? (Score:2, Interesting)
I've had mixed reliability results from both Seagate and Maxtor. Hopefully this union will take the best from both and result in
I'd rather see a reliability comparison (Score:3, Interesting)
Of the four I've bought in the last year and a half, two have failed. I've already replaced one and need to send the other back for a warranty replacement.
Failure seem high on those SATA drives that other people I know have, too.
Re: (Score:1)
Re: (Score:2)
Re: (Score:3, Insightful)
Mechanically speaking, the SATA connector doesn't seem particularly robust. I've had problems at work with one system in which the drive would occasionally disconnect and reconnect. Since the connectors use flat contacts that slide past each other and don't have much (if any) spring force behind them, it seems to me that you don't get as solid a connection as
Re: (Score:2)
You can generally get 5.25" SATA hotplug backplanes for about $30/disk. Common sizes are 3 trays in (2) 5.25" bays or 4 trays in (3) 5.25" bays. (There's even a 5:3 design but the inside of your 5.25" bays need to be clear of any obstructions along the sides.)
Re: (Score:2)
I have a RAID5 array of 4 Samsung SP1213C (120 GB) SATA drives on one of my systems, which has been truckin' along for two-and-a-half years of 24/7 operation without a hiccup. These are running off generic SiI PCI adapters, with Linux software RAID. No errors logged in the SMART error logs, either.
I've since put SP2504C single drives (250 GB) in a couple of my desktop boxen. They're too new (oldest=5 months) to assess long-term reliability, but I've had no tr
Re: (Score:2)
I knew someone who had a new drive go bad, and Samsung replaced it with no trouble at all after almost a year. They even sent him a slightly bigger drive (80GB instead of the 60GB that had broken).
So I've been pretty impressed with them in term
Re: (Score:2)
Be sure to try replacing the cable before deciding the drive is toast, even after testing. Those SATA cables are finicky.
Re: (Score:1)
Re: (Score:2)
Re: (Score:2)
Most of the recent complaints I've heard about disks that were either DOA or died after a short time were larger than 160 GB. 160 GB and less seem to have far fewer complaints.
Re: (Score:1)
I don't know what they did on the sata models , but they are junk.
Re: (Score:2)
Noisy, none less than 50dBA? (Score:1)
Granted, they are being measured without a case, and closer than the 1 meter that is conventional for such things. And I could accept that all drives are within this range of noise. But it still stinks for making audio equipment that can be used in the same space as a sensitive microphone.
10000 RPM SATA Drives? (Score:5, Interesting)
As of a year ago, Western Digital was the only one in the market. We need more competition for this so we can get cheap fast hard drives. SCSI is too expensive.
Re: (Score:2, Interesting)
They are there but speed difference isn't that big with fastest 7200RPM drives. Look SpintpointT or WD Caviar RE2 scores. Those are 7200, Raptors are 10000.
All in all, it looks like Western Digital and Samsung can and do make fast SATA drives. I remember some speculation that WD don't have SCSI markets/drives and so it won't bite its own sales if it makes its SATA drives run as fast as possible. Implying that other hard drive makers keep their SATA drives intentionally slower so they can keep selling SCSI
15000 RPM SATA Drives? (Score:2)
Where are the 15000 RPM SATA hard drives? 10K is too slow.
Seagate. The end. (Score:3, Interesting)
Re: (Score:1)
Re: (Score:1)
While we both have experienced abnormal failure rates for our respective drives, I certainly look at Seagate's warranty as a good thing, not bad. WD, Maxtor and Hitachi all dropped their warr
What about temperature? (Score:1)
My Favourite: Barracuda 7200.9 (Score:2, Informative)
When shopping for an additional drive for my linux box a couple of months ago I went with a Barracuda 7200.9 because of its low noise and low power consumption. At the time I was comparing the .9 with the .10 and found that the .10 had a 30-40% higher power consumption at the same capacity.
I'm still amazed that the newer drive consumes much more power (and runs hotter in consequence) with not much benefits at the same capacity.
Markus
Load times? (Score:2)
Our SATAs, our dual cores, our GB of DDR2, our PCI-E and it still takes almost a minute to boot up Windows.
No wonder I dont want to upgrade, progress isn't making things any faster.
Bill Watkins has been perfectly clear. (Score:2)
The old Maxtor products are going away. Period. Seagate has no intention to ship these products any longer than they contractually have to.
The Maxtor brand on the other hand, may continue on for some time, but it will only be a label on Seagate products. Maxtor had greater brand equity among consumers than Seagate does, and there is no reason to throw away that brand reco
Re: (Score:2, Interesting)
As for the notion that they have already reviewed all of these drives? They haven't. Not even in the 'roundup' review.
Re: (Score:2)
Re: (Score:1)
Does anyone have detailed stats of the new Microsoft operating system?
Re: (Score:2, Funny)
I can fit Vista on zero MB (Score:2)
Re: (Score:2)
Re:Not too happy with SATA in general (Score:5, Interesting)
Over 400 g5's deployed at my site over the last 2 years, all with SATA drives. There has been only one drive failure so far,
and it was premature so it was probably a bad unit or the movers roughed up his computer when he changed
offices, because it coincided with that event. We had a bunch of the last edition of the grey g4's with maxtor
hard drives that all seemed to fail within a few months of each other a 3 years back. We also
had a run of bad Maxtors in a batch of small form factor Dell desktops. We lost about 10 in one month.
Hooray for 3 year service plans !
the "Super"drives are dropping like flies though...I keep a stock of replacements in the closet. Good thing
a replacement with dual layer and lightscribe is around 60 bucks these days.
This is all highly subjective and anecdotal and not meant to slander maxtor in any way
Re: (Score:2)
As for the superdrives, they too in my experience seem to go wrong within 12-18months, generally on the burning side. As you say thank God they're cheap now and they're easy to get at.
Re:Not too happy with SATA in general (Score:4, Insightful)
I believe this has to do with the much weightier (relative to single-purpose devices) laser head arrangement that has the different types of lasers for each standard.
It also has to do with what you use the drive for. Eg: a drive that only ever reads or writes whole discs will last a lot longer than one constantly being used for random accesses.
Re: (Score:2, Interesting)
I don't think I'll ever go back to Maxtor, I think a lot of their problem was they were ATA133 not 100 like Seagate, the key to using Maxtors seems to be using them externally - I guess USB2/1394 doesn't push
Re: (Score:1)
btw - recognise your name from your newsletters - I really love the newsletters you guys send out ^_^
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
I have 2 drives at work (Raptors) that are 30C idle and 32C loaded, but they're in a bay cooler with a 80mm fan that pulls in from the outside and blows directly across the drive. Reckon those drives will last for quite a few years...
Re: (Score:2)
Along with poor packaging by vendors and poor handling by couriers.
Luckily, all the vendors I've used recently have boxed each drive in a wrap-around foam cage, in an individual box. I've heard horror stories of vendors shipping drives in the anti-static bag, with a handful of 'peanuts' in a large box, or
Re: (Score:2)
Re: (Score:2)