KingGremlin

Hitachi first to announce 1TB drive

Recommended Posts

Reported availability this quarter. Time will tell. It doesn't appear you can link to pages within Hitachi's site, so here is the front page which has links all over about the drive.

http://www.hitachigst.com

Tasty, and I was just getting ready to order up several 750 giggers. Ha! For some reason I think back to my 1.2 gig seagate, my first drive beyond 540mb. Very nice is all I can say. I hope the 32MB buffer is indicative of general read/write performance in some way... its a lot of space for a potential power loss. Then again I think back to my last 5 platter drive and it was a Deathstar 75gb. Its a fun drive torn apart the spindle motor has a good amount of torque and you can cut pretty deeply into a platter with a file.

Don't ask about the latter.

Share this post


Link to post
Share on other sites

I just noticed something weird in the datasheet. I was creating a post at SPCR forum to tell the news that have already been told here.

Anyway, here's the post. I made some summing up of the facts as direct linking to HGST might not work.

____________________

Hitachi Deskstar 7K1000

HDS721075KLAT60, HDS721010KLAT60,

HDS721075KLA330 and HDS721010KLA330

Hitachi Cinemastar 7K1000

HCS721075KLAT60, HCS721010KLAT60,

HCS721075KLA330 and HCS721010KLA330

www.hitachigst.com

Some of the main features listed (as HGST web site doesn't support direct linking and the datasheet is no longer a HTML document but PDF):

750GB (4 platters) and 1000GB (5 platters).

SATA2 with 32MB cache. PATA/IDE with 8MB(?).

Perpendicular recording (first 3.5 inch Hitachi to have).

Thermal fly-height control (something that is also in T7K500).

Load/unload technology (something all Hitachis use).

APM, AAM, etc. features.

Better non-recoverable error-rate (1 in 10^15 instead of 10^14).

Higher non-op and op shock tolerances (300G and 70G).

Power consumption (watts): 750GB / 1000GB

Random read/write: 12.8 / 13.6

Silent read/write: 9.9 / 10.8

Active idle: 8.1 / 9.0

Unload idle: 6.4 / 6.9

Low-rpm idle: 4.3 / 4.5

Allowed temperature ranges:

5 to 60 °C operating, -40 to 70 °C non-operating

(note: T7K250 5...55 °C

and T7K500 0...60 °C)

A half step backwards? Luckily most of us don't use their drives in near zero temperatures and upper limit is thus more important.

Anyway, probably not the most quiet drive around, but it will be sold under CinemaStar brand as well.

I just noticed... the model numbers of PATA models in in numbers "60". Latter is reserved / not used, but that first "6" stands for 16MB cache. Datasheet tells it has 8MB.

Model code deciphering guide:

H = Hitachi

D = Deskstar, C = Cinemastar, T = Travelstar, E = Endurastar, M = Microdrive, U = Ultrastar

S = Standard, T = Two/three platter, C = compact(?) (in combination with T this means 1.8 inch form factor), E = enterprise(?) in combination with T means desktop variant of Travelstar (for example E7K100)

72 = 7200rpm, 54 = 5400rpm, and so on

10 = 1000GB (maximum capacity of series)

75 = 750GB, 10 = 1000GB (capacity of the variant in question)

K = Kurofune (factory nickname for architecture capable of mounting up to 5 platters), V = Vancouver (up to 3 platters), P = Pathfinder (only 1 platter such as 7K80 and 7K160)

L = Low-profile (all 3.5 inchers)

AT = PATA, A3 = SATA 3Gbit/s, SA = SATA 1.5Gbit/s

6 = 16MB, 3 = 32MB, 8 = 8MB, 2 = 2MB

Share this post


Link to post
Share on other sites

"Hopefully these drives are rated for 24/7 operation..."

There's no physical difference between "rated 24/7" and "normal" drives. There's typically just a longer factory testing and some marketing BS added to a normal drive series. The result: drive will have a lower infant mortality rate.

After using the drive for sevaral days or weeks the added benefit of choosing a "24/7" drive has already passed. By that time, you have done some "testing" of your own and the factory testing period will be less meaningful. To evaluate long-term reliability, the "1 million hour MTBF" has no meaning at all (and it will not run for 1 million hours either). It's quite ironic, isn't it?

Which drive is not capable of operating 24/7? AFAIK they all are. The "not 24/7 capable" has probably originated from 75GXP in which time IBM told the drives should be used only 8 hours a day. I believe this has nothing to do with drive's incapability to operate over 8 hours strait but mainly because such a limitation will limit operational hours to one third of 24/7 use. No doubt 75GXP continued to fail even in 8/5 or 8/6 use like it did in 24/7 use. The number of failures would just be three times smaller.

Or actually there is a slight possibility that 75GXP might have been incapable to operating "over 8 hours a day". It has nothing to do with hours of operating, but instead hours of idling. With 8 hours a day, it'd mean it's powered on only when used - no significant idle periods at all. After the days work has been done: power off. No idling. Leaving it on over night means 16 hours of consecutive idling with R/W heads idling on the same track. This problem was eventually fixed by issuing a firmware update to 75GXP series drives (and there was probably one for 60GXP too). 120GXP and newer IBMs and Hitachis following the acquisition have had the reliability enhancing firmware since leaving the factory. No problems with idling periods: thus 24/7 use is OK with all current Hitachis. And very very likely with all non-Hitachis as well.

So if you meant to state your concern about the number of drives failing during first few weeks of integration to computers and storage arrays, I agree it is a valid concern for corporations as they certainly would like to minimize downtime. If you meant long-term reliability, you couldn't be more mistaken. Sorry.

Share this post


Link to post
Share on other sites

can't wait to see this and the seagate 1TB; and see how they compare. Hopefully the noise/seek levels won't be too high :X

Share this post


Link to post
Share on other sites
There's no physical difference between "rated 24/7" and "normal" drives.
That may be true, but that was part of the fiasco earlier for the IBM Deskstar 75GXP/60GXP era drives as you noted.

Hence, paranoia still exists. :D

Share this post


Link to post
Share on other sites
SATA2 with 32MB cache. PATA/IDE with 8MB(?).

The media transfer rate is listed as 1070 Mb/sec, that means, the disk can theoretically deliver data to the controller @133,75 MB/sec. Now, of course, real-world performance will only be around or little more than 100 MB/sec - but still faster compared to the Samsung T166 or WD Raptor (80-90 MB/sec), and defintely fast enough for the usual ATA (100 or 133 MB/sec) interface. So a real big buffer is only of limited use for ATA interfaces (seek times).

Uwe

Share this post


Link to post
Share on other sites

There's no physical difference between "rated 24/7" and "normal" drives.

That may be true, but that was part of the fiasco earlier for the IBM Deskstar 75GXP/60GXP era drives as you noted.

Hence, paranoia still exists. :D

Yeah, that was the point of my question, to get rid of the paranoia. I haven't touched a Hitachi/IBM drive in 5 years because of the 75GXP. :)

Share this post


Link to post
Share on other sites

hi

I wonder what external cases you will need to let this thing fly, when it already uses 8-9W idling (or powering up, which seems still to be unknown)?

regards

nicola

Share this post


Link to post
Share on other sites

Anything with a fan would be OK. Though one with a bigger fan is likely to be less noisy and fan bearings might last longer. You can even build an enclosure of your own if you have too much spare time.

9W is not too much for the external enclosures power adapter. It can handle such a load with no problems. And don't worry about start-up currents: 7K400 (Kurofune I) and 7K500 (Kurofune II) have the lowest start-up power draw of all drives benchmarked by SR. These drives use 15.0 and 15.2 watts (respectively) from 12V line. 7K1000 is Kurofune III.

Here's the benchmarks of Deskstars reviewed since SR started using Testbed 4. http://www.storagereview.com/php/benchmark...23&devCnt=6

T7K500 has slightly higher current draw from 12V line (19.6 watts or 1.63 amperes) during spin-up but still below average. External power bricks should be able to handle at least 2A. (Some are rated 2A 5V + 2A 12V and some are rated 3A 12V (with 5V regulator built inside enclosure so that) (thus 3A is shared between 5V and 12V). Or some of them may be rated for 1.5A 5V + 1.5A 12V but I would guess those are sustained currents.

Seagate Barracuda ES 750GB (which is pretty much identical to 7200.10 750GB) is the thirstiest of all drives benchmarked: 29.6 watts (2.47 amperes) on 12V line during spin-up. Not even SCSI drives rate that high. Other almost as thirsty drives include:

Maxtor MaXLine Pro 500GB

Seagate NL35.2 500GB

...and pretty much all other Seagates for that matter.

Where as Barracuda ES had highest peak power usage from 12V rail, ready-state continuous power usage is a completely different thing: SCSI drives use more (for quite obvious reasons).

Still: Barracuda ES and NL35.2 use most power (=generate most heat) of all 7200rpm drives. 9.4 watts on idle while those 5-platter Hitachis use "only" 8.5 watts.

I ain't saying Kurofunes are cool-running. I own a 7K400 and I would not dare to run it inside any passively cooled enclosure. And I have no reason to believe 7K1000 runs any cooler (or at least not the 1000GB variant).

Share this post


Link to post
Share on other sites
Guest 888

SATA2 with 32MB cache. PATA/IDE with 8MB(?).

The media transfer rate is listed as 1070 Mb/sec, that means, the disk can theoretically deliver data to the controller @133,75 MB/sec.

But... wasn't that internal media transfer rate given with control/ecc bits inluded? If yes then we must divide 1070 Mb/sec with at least 10(?) to get a real data transfer rate from platters. Then it will be just 107 MB/s or less for useful data - which is pretty realistic number if to compare with other current drives on market. Still I tend to think in practise its STR will be just below 100 MB/s... but it's just my speculation, nothin sure yet.

Also, I think this 8MB cache is a typo there on pdf-datasheet. As the model code just indicates 16MB to be.

Anyway, no more surprises for me - this model was long awaited and its 5x200GB construction was also foreseen already earlier then.

Share this post


Link to post
Share on other sites
Reported availability this quarter. Time will tell. It doesn't appear you can link to pages within Hitachi's site, so here is the front page which has links all over about the drive.

http://www.hitachigst.com

The CES 2007 Hitachi 1TB drive announcement got my attention. Didn't think they would be the first to make that claim, but considering Seagate's press release, will see who ships one first. Also, the announced online/discount price of $350 -$400 makes it even more interesting. Reminds me of the first 1GB drives coming to market over a decade ago, or at least back to the time when they were becoming somewhat economical and starting to sell in large volumes.

Share this post


Link to post
Share on other sites

... its a lot of space for a potential power loss. Then again I think back to my last 5 platter drive and it was a Deathstar 75gb. Its a fun drive torn apart the spindle motor has a good amount of torque and you can cut pretty deeply into a platter with a file.

Don't ask about the latter.

There's no physical difference between "rated 24/7" and "normal" drives.

That may be true, but that was part of the fiasco earlier for the IBM Deskstar 75GXP/60GXP era drives as you noted.

Hence, paranoia still exists. :D

Who would have ever thought the makers of the "Deathstar" 75GXP would ever be able to make another HDD that people might want to buy - especially considering the class action lawsuit filed against them for data loss and IBM's less then confidence building approach to sell the whole division to Hitachi shortly thereafter.

The "Deathstar" 75GXP was awesome for its time tho. Too bad so many of them croaked. I picked one up for a relative to replace a previous generation Maxtor drive which, incidently, didn't even last a year before I had to RMA for a replacement. I remember that 75GXP being so wicked fast and unbelievably quite. It was very impressive. I guess I was one of the lucky ones, it lasted about 3 1/2 years for him but I wound up finally replacing it for good measure because of growing complaints of problems on that system. Didn't think it prudent to keep pushing the luck with it

Edited by sam

Share this post


Link to post
Share on other sites
Guest 888
Also, the announced online/discount price of $350 -$400 makes it even more interesting.

However, Seagate's president in an interview not many weeks ago said that their upcoming 1T drive will cost about $700.

Yeah, I am pretty sure now Seagate's marketing people are pulling their hair... Still I think their drive will be more expensive than Hitachi's.

Share this post


Link to post
Share on other sites

sam: "Didn't think they would be the first to make that claim, but considering Seagate's press release, will see who ships one first."

Actually, wasn't it the other way around? Seagate made a less official annoucement 1 hour before Hitachi's official press release, thus Seagate made the claim first.

But claimed release dates were 1st quarter (Hitachi) and 2nd quarter Seagate. Unless Hitachi's release is delayed by more than 3 months from it's claim, Hitachi will be the first.

sam: "Also, the announced online/discount price of $350 -$400 makes it even more interesting."

Haha! :D

Seagate promises an online discount for 1TB drives? Discount sale of their HDD at a price of 350-400 dollars when Hitachi started selling 1TB drives 3 months earlier at a suggested (non-discount) retail price of 399 dollars.

888: "Yeah, I am pretty sure now Seagate's marketing people are pulling their hair..."

Yeah, no doubt about that. They have to actually offer "discounts" to match the standard pricing of Hitachi. And if the prices are to drop during the first 3 months, the announced "discounts" might still not be enough.

I consider it a good thing that Seagate can't monopolize the market of big HDDs any longer.

__

Ah, Deathstar ranting... I kinda assumed it'd come to this. :D

sam: "Who would have ever thought the makers of the "Deathstar" 75GXP would ever be able to make another HDD that people might want to buy"

Answer: ancient history (in terms of history of computing).

sam: "especially considering the class action lawsuit filed against them for data loss"

And WD had to settle and pay damages & legal fees because of a lawsuit claiming WD was misleading their customers by using SI-prefixes. Surely all of manufacturer's "mislead" in this way, and it just tells more about the judicial system of United States of Ameeeericaaa than the defendant of such lawsuits. All do the same, all screw up, only some are tried in court (or forced to settle under pressure).

There is no "justice" in judicial system, so I simply do not care. It's already enough that they have to actually pay money to ambulance chasers so they don't have to lose customers because of the same snake tongued people.

(Well, in case of IBM, boycotting them was justified. Not because of the failures but because of the way the scandal was handled. But that was IBM's decision. Different people, different times. Prehistoric.)

sam: "and IBM's less then confidence building approach to sell the whole division to Hitachi shortly thereafter."

What does that tell about Hitachi's confidence as they actually bought a division from a company that screwed their reputation?

Share this post


Link to post
Share on other sites

Hitachi probably thought they were getting a good deal.

Picture it this way. IBM practically invented the hard drive, and their hard drive division was leading or a major player in the market for a number of decades. Following one blip on the timeline, public opinon dropped. That would probably the best time to buy, as share prices will be low and therefore Hitachi would have gained IBM's hard drive division - which is one hell of an asset - for a significanlty lower price than they would have had to pay any other time.

It has been proven since that the supposed reliability issue ended with the 75GXP's, or depending on which skewed statistics you believe, the 60GXPs immediately following it. Every major HDD manufacturer has had at least one "bad apple" in their range. While IBM's was the most recent, it doesn't mean in any way that there is anything fundamentally wrong with them or their technology.

So you mention a class-action lawsuit. These happen all the time for the most mundane of things. Sony was served with several over the release of some music CD's. Does that mean all Sony products are bad? No. One team made one mistake, as did IBM. It's since been fixed, . I'm sure in a few years all this bad rep will be gone and you'll all be going on about the same crap over Seagate or WD or Fujitsu or Toshiba or someone else.

IBM/Hitachi have more experience than anyone else in implementing 5-platter hard drives, so much so that they would be "confident" in doing so while others aren't. All this paranoia about 5-platter drives... Have the 7K400 or 7K500 featured unusually high failure rates?

Share this post


Link to post
Share on other sites
But claimed release dates were 1st quarter (Hitachi) and 2nd quarter Seagate. Unless Hitachi's release is delayed by more than 3 months from it's claim, Hitachi will be the first.

Q1 could be March 31, Q2 could be April 1. So a delay of a ms could already turn the tables.

Share this post


Link to post
Share on other sites

"One team made one mistake, as did IBM."

And like with IBM disaster, Sony BGM disaster wasn't that big to begin with but got bigger due to bad handling of the situation.

Sony's initial problem was buying services from a non-experienced supplier of software. There was some major flaws in the original product:

- end-user was not informed of all the alterations software did. (Only that some special player would be needed.)

- hiding the DRM software in directory hidden by using a rootkit. Also allowed any malicious programs and viruses to cloak themselves.

- no uninstall was supplied.

When they finally reacted, they merely offered to uncloak the DRM, not remove it (even though customer was not informed of some system alterations that happened during installation of Sony DRM). Also, this uncloaking update included an ActiveX component which should have been only used during download phase but was left in active state and which created a bigger security hole (as malicious programs could install themselves from any website without asking permission). According to official truth a real uninstaller could be obtained also, but not by direct download. I don't remember for sure if anyone actually received the real uninstaller despite numerous emails sent.

Kinda resembles the GXP scandal, doesn't it? Nothing special at first (= a normal screw-up). No immediate fix, refusal to admit the problems to public, bad customer service, etc., and till the very end: no regret => a BIG problem.

"Q1 could be March 31, Q2 could be April 1. So a delay of a ms could already turn the tables."

Yeah, I know. If Hitachi misses one day off March 31, the actual release is early Q2. But remember: if Seagate misses June 30, the actual release is Q3.

The difference between releases is 3 months on "average". It might be less (if Hitachi fails to meet their claim or if Seagate is quicker than what they claimed) but it may as likely be that is more than 3 months.

Share this post


Link to post
Share on other sites

I'm bumping this thread to ask if there's any new information on release dates.

I could have bumped the "Where, oh where, are the Terabyte drives, Due out in the next 3 months....." thread but I think this thread is more "on topic" than that one. Let's keep it that way. If you want to comment on Seagate's upcoming 1TB drive or speculate about differences between Seagate's 1TB and Hitachi's 1TB (including factors that affect reliability) post in:

http://forums.storagereview.net/index.php?...=24063&st=0

...or start a new thread..

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now