Samsung Spinpoint F3EG HD203WI 2TB

[NT]

Thanks.
I'm not interested in desktop/consumer grades; only looking at near-lines.
2TB just has too much of a unjustifiable premium atm. (let alone near-line)

I'm delaying for a few more months, in the meantime I'm going to make do with just one 1TB WD Black as boot/storage volume for HEPC*.
I have plans for a dedicated native VMM/NAS in a few months, at that time I will reinvestigate 1.5-->2TB/near-lines.

*may also get one cheap barefoot controller based 30GiB SSD.

[soujir0u]

I just bought a few of the 203WIs.

Their idle noise is considerably higher than the 1TB F1 and F3 drives I have.

[filip007]

What about Seagate 2TB XT?

I have second thoughts about HD Tune and SMART reading it make my Samsung F1 hotter and loader and same go to my another HD161GJ Samsung drive.

I have put some bubble bag with foam around it i hate that HDD read noise...

[Quindor]

I have been doing some research into which drives might be compatible with my Adaptec 5805 raid adapter (not really suited for silent pc's because of the enourmous heat it generates).

In this search and my experience over the last years. RAID type disks do not exist for nothing. But, doing closer research this is only because of one single setting : TLER / CCTL / etc.

It's quite simple.

Action to take when reading a sector is "hard"
Desktop disk - try unlimited time
RAID disk - try 7 seconds max, then report back

What does the controller expect
Desktop controller - Wait for the disk to respond, hangup whole system if needed
RAID controller - Wait max 8 seconds (often default value) then declare disk bad and evict/fail disk from raid set

As you can see, mixing a desktop disk with a raid controller can produce unwanted results. Disks will be dropped, while nothing is really wrong with it, etc. In the end, it will become allmost impossible to rebuild the RAID array anymore.

Good, that's the background. I'm looking to purchase 7 of the Samsung HD203WI drives for my new RAID array. But, first I need to know if it's possible to use smartctl and change the value of the TLER of the disks. If so, they are very likely safe to use in a RAID array over time (problems often don't manifest immediatly but after usage and wear)

If someone could test the following on their disk. Your controller does not matter (RAID controllers are bit more tricky and need a different approach though).

Windows users, download a tool called smartmontools. That gives you the smartctl executable, which we'll need. You can download it here : http://smartmontools-win32.dyndns.org/s ... -setup.exe . For other moderns OS's flavors, please find look on the website : http://smartmontools.sourceforge.net . You will have to build the SVN version to be able to use this specific command.

Smartctl can be used for many other things besides this. It can tell you much about the life of your disk or say the current temperature for monitoring, etc. etc.

None of these tests below will touch your data.

When you have that installed, say on a windows machine and your samsung HD203WI is your "d:" drive, execute the following :

smartctl -l scterc d:
If correct, this will give you the following feedback on a desktop disk
SCT Error Recovery Control:
Read: Disabled
Write: Disabled

Now we are going to try and change that value :

smartctl -l scterc d:
If that works, you will see the following feedback :
Read: 70 (7.0 seconds)
Write: 70 (7.0 seconds)

To put it back to original values again, either turn off the power of your system, or run "smartctl -l scterc,0,0 d:"

When running a RAID array. This command will have to be run for all your disks every boot, because the value resets itself after the power is turned off. But since that is easily scriptable, you can fix this problem, before it occurs!

[MoJo]

Anyone find power consumption measurements for these drives?

No-one seems to have measured them yet.

[Quindor]

I have moved my questions and topic over to the storagereview.com forums. I have tested the disks and they support the TLER error control.

[sireddie]

CCTL works on these drives!

[sireddie]

wait for it...

[sireddie]

...almost there

[sireddie]

And after 3 posts, here is the URL:

http://forums.storagereview.com/index.p ... tler-cctl/

[JazzJackRabbit]

Wonderful news on TLER control. If my RAID5 array runs out of space, these are the drives I'm going to buy, that is unless 3TB drives do not come out first

[whiic]

F1 - 7200rpm spindle, 333GB/platter
F2EG - 5400rpm spindle, 333GB/platter
F3 - 7200rpm spindle, 500GB/platter
F3EG - 5400rpm spindle, 500GB/platter

I think that list is both incomplete and partially erroneous. First, it's missing F1EG. And F2EG is probably 500GB/platter in all capacities (not just 1.5TB which was pointed out that it wouldn't fit into 333GB/pl theory).

There was F1EG and it had 5400rpm & 333GB/pl. F2EG practically had same specs (aside from rpm) as F3. F3EG seems more or less to be a renamed F2EG... or it might have the same rpm and 500GB platters but differ from F2EG in some other aspect. For example:


Note that in the picture the HDD is equipped with load/unload mechanism found in IBM/Hitachi Deskstar and WD Greenpower/Caviar Green lines of HDDs (and in all laptop HDDs). Previous Samsung HDDs did not have unload ramps but were contact start/stop type with landing zone for sliders located on inner diameter of platters.

I also ain't sure about the 4-platter theory. 2TB's model number is HD203WI where as 1.5TB of F3EG is HD153WI. "W" is a letter that denotes a combination of rpm and number of head. It could of course be that F3EG 1.5TB has lower density platters (4 of them) than 2.0TB, but that would mean that F3EG 1.5TB would be slower than F2EG 1.5TB!

F2EG is model numbered as HD154UI... "U", not "W". Thus it would make sense if F2EG had 3 500GB platters and F3EG had either lower-density or short-stroked platters. F3EG 1.5TB could possible be an outlet for 2.0TB drives that fail validation at maximum capacity...

G: 5400rpm, 1 heads (1 platter) (F3EG: HD253GI)
H: 5400rpm, 2 heads (1 platter) (F3EG: HD503HI; F2EG: HD502HI)
J: 5400rpm, 3 heads (2 platter) (F3EG: HD754JI)
S: 5400rpm, 4 heads (2 platter) (F3EG: HD105SI; F2EG: HD102SI, HD103SI)
U: 5400rpm, 6 heads (3 platter) (F2EG: HD153UI, HD154UI)
W: 5400rpm, 8 heads (4 platter) (F3EG: HD203WI, HD153WI)

Note the one emphasized with red text: it's the only difference one that does not match 500GB/pl within F2EG and F3EG. I could make a more comprehensive list with 7200rpm Samsungs as well but I won't bother, not related to topic.

There's a missing letter between S and U and between U and W but that's only what is to be expected as number of heads are 4, 6 and 8 respectively. There's room for 5 and 7 head HDDs if they wanted to make them.

[__Miguel_]

F1 - 7200rpm spindle, 333GB/platter
F2EG - 5400rpm spindle, 333GB/platter
F3 - 7200rpm spindle, 500GB/platter
F3EG - 5400rpm spindle, 500GB/platter

I think that list is both incomplete and partially erroneous.

That list was put together with the information I had at the time, and I never claimed it to be 100% accurate (as already pointed out earlier).

Anyway, thanks for the input, it shows you took your time with this one.

Note the one emphasized with red text: it's the only difference one that does not match 500GB/pl within F2EG and F3EG. I could make a more comprehensive list with 7200rpm Samsungs as well but I won't bother, not related to topic.

Yes, there always seems to be one "odd one out" in any chart... I really wished Samsung (and WD, while we're at it) published its drive naming scheme (like Seagate), so it would be easy to know what to expect just by reading the label...

I tell you, Foxconn naming schemes for mobos are GREAT in that regard (and, to some extent, Gigabyte, though it's not easy to pinpoint extra features, like you can with Foxconn).

I think that comprehensive list of both 5400rpm and 7200rpm Samsung HDDs would be a great idea. I'd vote for sticky in a heartbeat!

Cheers.

Miguel

[Michael Sandstrom]

I've had an HD203wi for a couple of weeks and idle noise seems slightly higher than the HD103UJ and HD103SJ drives. None of the ten Samsung SATA drives I've used in the last 5 years have exhibited noise or reliability issues. Samsungs can often be found at very competitive prices. I don't think the mass market has discovered how good Samsungs are.

[__Miguel_]

I've had an HD203wi for a couple of weeks and idle noise seems slightly higher than the HD103UJ and HD103SJ drives.

Twice the platters, according to whiic's scheme, so that is to be expected.

None of the ten Samsung SATA drives I've used in the last 5 years have exhibited noise or reliability issues. Samsungs can often be found at very competitive prices. I don't think the mass market has discovered how good Samsungs are.

I hear you on that one.

Apart from a "fluke" I had with a 3-month-old HD154UI, which suddenly developed 600+ pending sectors and made WHS complain about bad sectors, write errors and CRC errors, the longest running HDD I had was a 20GB IDE Samsung drive. It finally gave out (my guess is it was the controller, since I still can't detect head crashing sound) about one year ago, after being in service for over I don't even know how many years.

I believe that drive was the one that came with my first "real" PC: a Pentium Pro 180 machine (yes, THAT old).

While that drive was noisy (back then Maxtor was the way to go for low-noise drives), I could never complain about it. And newer Samsungs are great price/noise/performance drives (especially price, I can get a 2TB Samsung drive for about €50 LESS than a comparable WD drive...).

Cheers.

Miguel

[apicus]

i am having problems with the SAMSUNG 2TB HD203WI in my 4 bay NV+
expanding to these disks does not seem to work, not even force expansion, i am stuck with the old setup.
all netgear says, like a parrot, is "it is not in the compability list"...

[JamieG]

I just picked up one of the HD203WI drives as well for HTPC storage / backup purposes.

I'm in the process of copying my video library to it now, but I just thought I'd add some comments on the noise levels.

The HD203WI is noticeably quieter at idle sitting on some foam on my desk when compared to the F1 1Tb drive I have in an external HDD cage. Admittedly the F1 is hard-mounted to the cage but there is a significant difference in noise level.

On the other hand, the idle noise of my Seagate FreeAgent Go 500Gb USB external HDD (I think the HDD inside is a 2 platter 5,400rpm Seagate Momentus 5400.x model) is quieter than the HD203WI at idle.

There isn't much of an increase in noise when I've been copying files to the HD203WI, just some occaisional muted seeking noise. Copy speeds from temporary storage on my FreeAgent Go to the HD203WI is about 34-35MB/s.

Anyway, since the HDD is going into my HTPC, which isn't the quietest computer up close but is fine for watching movies from the couch, I'm pretty happy with the sound level for the size of the storage.

[__Miguel_]

The HD203WI is noticeably quieter at idle sitting on some foam on my desk when compared to the F1 1Tb drive I have in an external HDD cage. Admittedly the F1 is hard-mounted to the cage but there is a significant difference in noise level.

Well, I think that is to be expected. The 1TB F1 is a 7200rpm HDD made to be as fast as possible (Samsung really wanted the speed crown when the F1 launched), which combined with hard-mounting and enclosure resonance is kind of a recipe to loudness...

On the other hand, the idle noise of my Seagate FreeAgent Go 500Gb USB external HDD (I think the HDD inside is a 2 platter 5,400rpm Seagate Momentus 5400.x model) is quieter than the HD203WI at idle.

That is also to be expected. The Momentus is a 2.5'' HDD, and even 7200rpm ones are usually quieter than 5400rpm 3.5'' HDDs. That's why SPCR has been recommending 2.5'' HDDs as system drives for years. And in this instance, the fact that you're dealing with a 4-platter design doesn't really help with the comparison...

Copy speeds from temporary storage on my FreeAgent Go to the HD203WI is about 34-35MB/s.

That's a rather bad test for throughput. The FreeAgent Go is USB-based, which means you will be (and in fact are) limited by the interface. 35MBps is the maximum you can squeeze out of USB 2.0. That is roughly 1/3 (probably slightly more than 1/3) of the maximum speed the F3EG is capable of handling, so it's not even breaking a sweat when you're transferring stuff like that.

If you want maximum noise output, try running a benchmark tool in a random read pattern for a few minutes. That will probably give you the maximum noise output you can expect from the HDD; real life will probably be a lot lower, random reads are not really this type of HDD's normal usage pattern.

Cheers.

Miguel

[JamieG]

Yep, I know all that Miguel.

Nevertheless, thought I'd share my thoughts.

[mLy!]

Is there any new info on the differances between F3EG and F2EG 1.5TB?

the F3 seems a bit more expensive (5€) in the shops that have them

[__Miguel_]

mLy!, check whiic's post from April on this very thread. It has a comprehensive breakdown of the Samsung FxEG line of HDDs.

I personally haven't seen any reviews of the 1.5TB F3EG, so I can't say if it's a re-badged F2EG, a short-stroked 2TB F3EG or something in-between.

Common sense and good engineering/marketing practices dictate that the 1.5TB F3EG *should* be faster than the F2EG version, but right now only a Samsung rep or an independent review could shed some clear light on this one.

If whiic's analysis is correct, though, you could experience lower operating temperatures, noise and power draw with the 1.5TB F2EG, since it's only a 3-platter design (versus 4 platters for the F3EG). Not a very large difference, but it might matter for your usage model.

Cheers.

Miguel