dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
774
TomBrooklyn
Premium Member
join:2002-03-04
Brooklyn, NY

TomBrooklyn

Premium Member

Seagate Barracuda: Is It Faster Than Average?

Are Seagate Barracuda hard drives faster than the typical 7200 rpm hard drive?

koitsu
MVM
join:2002-07-16
Mountain View, CA

koitsu

MVM

Most Seagate Barracuda series drives are 7200rpm!

Can you be a little more specific?
TomBrooklyn
Premium Member
join:2002-03-04
Brooklyn, NY

TomBrooklyn

Premium Member

said by koitsu:

Most Seagate Barracuda series drives are 7200rpm! Can you be a little more specific?

I know the Barracuda is 7200 rpm.

How's this: Are Seagate Barracuda hard drives faster than other typical 7200 rpm hard drives?

If I'm not mistaken, Seagate markets, or at one time marketed, the Barracuda drive as being faster than other hard drives.
JoelC707
Premium Member
join:2002-07-09
Lanett, AL

JoelC707

Premium Member

According to their wiki page, the Barracuda was the first commercial drive to hit 7200 rpm. IIRC I remember something about the Barracuda ATA IV or so was "faster" or whatever than the current drives at the time but that was a long time ago and I honestly don't remember. That could simply have been benchmark tests or something. I do remember Seagate advertised a "fluid dynamic bearing" with those models. I doubt that had anything to do with it but who knows.

koitsu
MVM
join:2002-07-16
Mountain View, CA
Humax BGW320-500

koitsu to TomBrooklyn

MVM

to TomBrooklyn
You mean real-world performance? No, they're no better or worse than other brands. What you've read is pure, 100%, marketing. There is no speed advantage to Seagate drives over any other brand at this point in time.

One thing I will point out though: over the past 4-5 years many of the specifications documents for MHDD products, which used to be fairly detailed, are now lacking key pieces of information (seek time is becoming harder and harder to find, head or platter count is rarely mentioned any more, etc.). Marketing is also responsible for this. Really, really angers me.

pnjunction
Teksavvy Extreme
Premium Member
join:2008-01-24
Toronto, ON

pnjunction to TomBrooklyn

Premium Member

to TomBrooklyn
When Seagate launched their 1TB/platter 7200rpm drives (product codes STx000DM001) they were notably faster than the Hitachi or WD units, about 30% faster on large sequential read/write and also faster on smaller operations. This is because higher density generally translates to more speed.

»www.hardwarecanucks.com/ ··· d-4.html

»www.hardwarecanucks.com/ ··· d-5.html

»www.hardwarecanucks.com/ ··· d-6.html

I'm not sure if Hitachi or WD has updated their offerings to keep pace, I haven't heard about it.

JimE
Premium Member
join:2003-06-11
Belleville, IL

JimE to TomBrooklyn

Premium Member

to TomBrooklyn
And I would like to point out that the times (ie: faster or slower) are in miliseconds. This is not something that you notice in day to day activities.

pnjunction
Teksavvy Extreme
Premium Member
join:2008-01-24
Toronto, ON

pnjunction

Premium Member

said by JimE:

And I would like to point out that the times (ie: faster or slower) are in miliseconds. This is not something that you notice in day to day activities.

Which times? IO or access times I agree, the difference won't be too noticeable.

On something like a large copy operation though, a faster drive will take proportionally less time. For example moving 1200MB at 120 MB/s will take 10s but at 150 MB/s it will only take 8. Not huge difference but nonetheless perceptible (especially for the impatient lol).

Of course there is still the question of whether the other end of the operation can feed/eat data that fast. If I copy from my 500 MB/s SSD to a drive I can see the full speed, but when I copy to/from my NAS it is limited by the LAN to about 100 MB/s regardless.

koitsu
MVM
join:2002-07-16
Mountain View, CA
Humax BGW320-500

koitsu

MVM


WD10EFRX, quick benchmark

WD10EFRX, SMART stats
I seriously question those benchmarks.

I've actually had one of those Seagate drives (well, the 2TB model, a.k.a. ST2000DM001), and the performance was nothing to write home about, especially when compounded by the fact that the drives, much like the WD Greens, excessively park their heads.

I have WD Caviar Black drives that match those sequential I/O numbers (160-170MBytes/second reads), and I have WD Caviar Black drives that do 100-110MBytes/sec as well (heavily used drives).

The "1TB density platter means faster I/O" claim is nonsense from everything I have tested in the field.

The only drives WD makes as of this writing which use 1TB platters are the WD Red series. Yes, that means the WD10EFRX is a single-platter drive (I have one, screenshots attached) -- very easy to tell from the weight.

pnjunction
Teksavvy Extreme
Premium Member
join:2008-01-24
Toronto, ON

1 edit

pnjunction

Premium Member

Indeed my ST3000DM001 did have the head parking problem in my NAS. Clickety click all the time. Although I think the drives would still last for years according to math on load cycle count, I fixed it with the hdparm command and they're settled down now. I didn't think that problem happened in windows I though it was just an incorrect hdparm level in my NAS? (It was cranked to the max power-saving)

You're showing an HD tune for a 1TB/platter drive, so I'd expect it to be closer to the 1TB/platter seagate although I haven't seen much on the performance of those drives.

I said higher density generally translates to higher speed, you're right it might not mean higher I/O if the controller is poorly done. It should however translate to higher sequential read/write unless something has gone wrong, that's fundamental to how these things work. (fixed speed of 7200rpm + fixed platter size in a 3.5" + higher density = higher data rate moving past the head). Higher density is the main reason sequential performance has kept increasing. (Meanwhile access time does not increase much, since the distance the head needs to travel is staying the same.)

Here is another review which shows the WD black at about 130MB/s and the seagate TB platter at about 160.

»www.storagereview.com/se ··· 000dm001

Tom's has about the same results for sequential tests but they do give the edge to WD in random performance.

»www.tomshardware.com/cha ··· 928.html
»www.tomshardware.com/cha ··· 929.html
»www.tomshardware.com/cha ··· 926.html
»www.tomshardware.com/cha ··· 926.html

In the end though it might come down to one thing I didn't realize before I bought my ST3000DM001 drives: they only have a 1-year warranty while the WD black has 5, the WD red has 3. You pay for it though in money/capacity. The WD black only goes up to 2TB which is 1.8x the cost of the Seagate 2TB, while the Red 3TB is 1.5x the price of the Seagate 3TB. If the Red was out when I bought 4x3TB for my NAS, I probably would've gotten it (it is limited by LAN anyways).

signmeuptoo94
Bless you Howie
Premium Member
join:2001-11-22
NanoParticle

signmeuptoo94 to TomBrooklyn

Premium Member

to TomBrooklyn
Have we reached a point where there is diminishing returns for greater areal density? Could that even happen?

sk1939
Premium Member
join:2010-10-23
Frederick, MD
ARRIS SB8200
Ubiquiti UDM-Pro
Juniper SRX320

sk1939

Premium Member

I don't think we have yet, but we will in the near future I predict. There is so much you can fit on a platter and still keep the 1/2 height 3.5" form factor. You could double the space per disk if companies would go back to full height drives, but I don't see that happening.

signmeuptoo94
Bless you Howie
Premium Member
join:2001-11-22
NanoParticle

signmeuptoo94

Premium Member

I'm not sure, but I wonder if you misunderstood my question. What I am asking is, is there a point of diminishing improvement in performance even with greater areal density of data, if, say, other factors are involved...

sk1939
Premium Member
join:2010-10-23
Frederick, MD

sk1939

Premium Member

Koitsu would know better, but again I predict yes, because you can only make the read/write heads so small before you start to lose efficiency due to the number of sectors.

koitsu
MVM
join:2002-07-16
Mountain View, CA
Humax BGW320-500

koitsu

MVM

The actual physics or mechanics behind platter density vs. performance are something I don't quite understand. That's a very cool/neat science (a blend of actual science combined with technological science) that I don't have familiarity with. I'm just a simple caveman...

signmeuptoo94
Bless you Howie
Premium Member
join:2001-11-22
NanoParticle

signmeuptoo94

Premium Member

Yeah, but even a caveman can do it...

koitsu
MVM
join:2002-07-16
Mountain View, CA
Humax BGW320-500

1 recommendation

koitsu

MVM

said by signmeuptoo94:

Yeah, but even a caveman can do it...

Square/rectangle platters?