backtop

Print E-mail del.icio.us 38 comment(s) - last by TheGreek.. on Jun 12 at 1:40 PM
Seagate's drives are dense: 180 Gb per square inch

Seagate sent out a press release announcing that it has begun worldwide volume shipments of the industry's highest areal density desktop hard drive. At 250GB-per-disc, 3.5-inch disc relies on Seagate’s second-generation perpendicular magnetic recording technology to achieve its high density.

Packing a data density of 180 Gbits per square inch, Seagate says its one-disc Barracuda hard drive sets new benchmarks for power consumption, acoustics and performance for its desktop PC hard drives product line. The Seagate Barracuda 7200.10 drive is built with an SATA 3Gb/s interface and will serve as the foundation for Seagate's 1-terabyte desktop, enterprise, consumer electronics and external hard drives.

"Seagate remains focused on leading the hard drive's pivotal transition to perpendicular recording technology and maintaining our areal density leadership in order to meet our customers' growing storage capacity and reliability needs," said Brian Dexheimer, chief sales and marketing officer for Seagate. "This product's leading areal density epitomizes our efforts to deliver technologies that are unmatched in allowing organizations and consumers to store, protect and share digital content."

Comments     Threshold


This article is over a month old, voting and posting comments is disabled
Which is better?
By Spivonious on 6/8/2007 10:37:29 AM , Rating: 2
More platters and thus more readheads, or one big platter? I would think that hard drive manufacturers could build in some RAID 0-like functionality and treat each platter as its own disk. Think of the performance gains.




RE: Which is better?
By Talcite on 6/8/2007 10:44:46 AM , Rating: 2
That would mean that the hard drive manufacturers would also have to fit in 5 motors in a single 3.5" HDD. I don't think that's going to happen any time soon.


RE: Which is better?
By Von Matrices on 6/8/2007 10:56:51 AM , Rating: 3
I'm not an expert, but I would think that the the OP's idea would be more like writing the data in the same coordinates on each side of each platter so that all the locations on the platters align vertically. Thus, only one motor is needed to align every head to those exact coordinates to read the data from every platter simultaneously.


RE: Which is better?
By Screwballl on 6/8/2007 11:02:17 AM , Rating: 2
why would 5 motors be needed... if it is setup with RAID then it would just copy the same information on multiple platters at the exact same time. If it detects a platter may be corrupt it can display a message that it is time to replace the hard drive and just use the backup on the non-corrupt platter. If it is continued to be used it just disables RAID.
Just make this an option so that we can use the single drive as either a 250GB with per-platter backup or as a single 500GB which would be the same as almost all other consumer hard drives today.


RE: Which is better?
By HrilL on 6/8/2007 11:22:16 AM , Rating: 2
I see your logic but most of the time it is not the platters that fail first It is a different part of the drive so this isn't really a good idea. But if it were raid 0 then I can see doing it for 2 times the performance. Or close to that. It would help them get data throughput closer to sata 3Gb/s max of 375MB/s of course that is the theoretical max but even a drive that has throughput of 200MB/s would be very awesome. But I think by doing this it would hurt their server drives sales so I don't see this happening any time soon. Since the fastest 15k drive from Seagate has something like 124MB/s of throughput and is only 300GBs.


RE: Which is better?
By kalgriffen on 6/8/2007 11:30:48 AM , Rating: 2
Wouldn't it be the same performance since the drive is still limited by the read/write head?


RE: Which is better?
By Ender17 on 6/8/2007 11:30:46 AM , Rating: 2
First of all, RAID0 offers very limited performance gains as can be seen by any of the many reviews AnandTech has done on the subject.

As far as single drive RAID0,
http://faq.storagereview.com/tiki-index.php?page=I...


RE: Which is better?
By Von Matrices on 6/8/2007 11:44:29 AM , Rating: 2
quote:
RAID0 offers very limited performance gains

I read the article, and yes, complexity of components is increased, but from the description, the performance increase of internal RAID 0 compared to traditional drives is drastic, even to the point of saturating SATA 150 and possibly even SATA 300. However, I still see the limited storage space of one drive being the main limitation to single, high-performance drive adoption by enthusiasts.


RE: Which is better?
By mindless1 on 6/9/2007 6:55:34 PM , Rating: 2
No.

IN isolated testing it shows it's strengths. In most tasks it is actually SLOWER than having both drives as independent volumes for concurrent I/O.


RE: Which is better?
By Sh0ckwave on 6/9/2007 2:51:58 AM , Rating: 3
I tried RAID0 with 2x 320GB seagate 7200.10 and a Gigabyte 965p-DS3 motherboard. Heres the results:

RAID0: 128 MB/s average read
1 Drive: 65 MB/s average read

Those old benchmarks they did years ago have no relevance anymore. RAID controllers and hard drives have come a long way since then.


RE: Which is better?
By RamboZZo on 6/8/2007 11:57:29 AM , Rating: 2
Seagate did make a drive a long time ago that had independently seeking heads for each platter. The technology is there to do all this stuff but its really a matter of cost,practicality, and reliability. Harddrive profits are very thin and implementing any of this will just make more expensive drives. People just want cheap drives. It also makes the actuators more prone to failure and thay is already on of the most delicate parts of a drive. From a practical standpoint its just better to get more drives.


RE: Which is better?
By Moose1309 on 6/8/2007 12:14:41 PM , Rating: 2
This has already been implied in previous posts but ...

Number of platters tends not to impact performance much. However the higher density does increase sustained read speed.

Also fewer platters creates quieter acoustics and lower thermals. Also Steve Gibson (spinrite guy) claims there's a study linking more platters to higher failure rate.

This is why I held off on the Hitachi, and am waiting for this 7200.11. A 500 GB 2-platter drive would be nice.


RE: Which is better?
By patentman on 6/8/2007 12:26:07 PM , Rating: 3
More platters does lead to higher failure rate, because you have more heads and thus more opportunities for a fatal head crash.

if anyone has questions about Seagates tech, pm me on the anandtech forums. I issued most of the patents Seagate owns re: perpendicular recording technology.


RE: Which is better?
By ninjit on 6/8/2007 4:12:35 PM , Rating: 2
Uh, that's essentially how multi-platter drives already work.

Let's say you have a 2 platter drive, the 2 heads are connected to the same mechanical arm (moved by a single motor).

To write out an 8-byte block, 4-bytes are written to one platter and 4 to the second simultaneously, in the same areal location on both.


RE: Which is better?
By sleeprae on 6/8/2007 7:28:29 PM , Rating: 2
Actually, the two writes occur sequentially, not simultaneously. Reference the StorageReview article posted.


RE: Which is better?
By Treckin on 6/8/07, Rating: 0
RE: Which is better?
By TheGreek on 6/12/2007 1:40:44 PM , Rating: 2
Why in the world would anyone downrate Treckin's post? What's wrong with it? Please explain.


RE: Which is better?
By gknoell on 6/8/2007 8:57:30 PM , Rating: 2
Actually more Platters seem to be used already in a RAID-0 like fashion.
I got his from the Technical spec sheet of the WD Caviar SE series:

Data transfer rate
Buffer to disk (multi-platter) 748 Mbits/s (maximum)
Buffer to disk (single platter) 475 Mbits/s (maximum)

This means that the Multi-Platter HDs run faster, doesnt it?


RE: Which is better?
By mindless1 on 6/9/2007 6:53:19 PM , Rating: 2
What an ignorant post.

The performance gains of raid are due to independent discs having independent cache, independent ability to secure bus I/O, indepedant internal transfer rates, etc.

The whole point is it is NOT in any way the same drive. What you propose is defintely worse, not better performance.


By PAPutzback on 6/8/2007 10:11:15 AM , Rating: 3
Is it just 4 platters IE. 4x250 = 1TB. I think it must be more because aren't there already TB drives out there. So the current one that aren't running 250 Gig platter must have something closer to 5 platters of 200. I was hoping to see 2TB drives by next fall for 3-400 bucks.




By techfuzz on 6/8/2007 10:22:16 AM , Rating: 1
Back in January, news was that Seagate would release a 1TB drive in about 6 months, which happens to be right about now. It's very conceivable that such a drive would have 5-6 platters and consume considerable power and generate a lot of heat.


By namechamps on 6/9/2007 8:19:04 PM , Rating: 2
Math wasn't your strong point was it?

This is an article of 250GB/platter drives. So a Seagate 1TB drive using 250GB platters would be how many platters? Answer is 4. Why is this important? The current 1TB drive from Hitachi uses 5 platters. Going from 5 to 4 platters REDUCED noise, heat and cost. It INCREASES reliability and performance.

Seagate generally doesn't make 5 platter drives. Their last flagship drive was 750GB on 4 platters = 187GB / platter. So Seagate has improved density by about 33% in about a year (750GB 7200.10 was released around April 2006).


By Martimus on 6/8/2007 10:25:22 AM , Rating: 2
Hitachi is using five 200GB platters for their Terabyte drive.

http://www.anandtech.com/storage/showdoc.aspx?i=29...

I don't think that there is a hard limit on the number of platters that can go in a hard drive, but if there are too many they won't fit in a standard sized 3.5" HDD case. I would guess that 5 platters is about all you can fit, or else we would have higher capacity drives now.


By HrilL on 6/8/2007 11:26:36 AM , Rating: 2
Yeah I think 5 is the max they can fit too. And as for Seagate they said they would have 4 platters for their 1TB drives. They are farther ahead of Hitachi in patter density technologies.


By hans007 on 6/8/2007 3:53:53 PM , Rating: 2
generally the reason you cant put more platters in is that the rotational mass is too much for the motor to rotate at the given rpm.

they could probably make a drive with 10 platters or something in a double height drive if they lowered the rpm which would make the drive slow.


By RamboZZo on 6/8/2007 11:38:13 PM , Rating: 2
That was common until not that long ago. Double height drives with massive storage capacities at the expense of speed. I think the last such drive Seagate made was the 140GB barracuda maybe 5 or so years ago. It had to run at 7200rpm instead of 10,000rpm. For the time it was a monstruosity in storage capacity.


By RamboZZo on 6/9/2007 12:53:26 AM , Rating: 2
http://www.storagereview.com/articles/200104/20010...
This is the drive I was refering to. It was a 180GB drive with TWELVE platters in the large 1.6" half height factor. The correct name for the double form factor is actually half height, not double height or full height. Standard 3.5" drives as we know are actually 1" low profile, not half height which is the large, no longer used size.


By Suomynona on 6/9/2007 3:56:09 AM , Rating: 2
The form factor for these has been around for a while. i remember pulling a half-height drive out of an old NEXT Cube I was modding. I think it was something like a whopping 800mb storage capacity.


By RamboZZo on 6/9/2007 9:38:48 AM , Rating: 2
Yeah I know. I used to have a whole buch of old Quantum models I think. They where about 1gb. I also had some ancient full height 5.25" drives. I regret throwing them out. Half heigh was very common on high capacity scsi and FC drives back then. That seagate model I refered to was one of the last of its kind that I know of. I don't think they ever used the form factor again.


By iVTec on 6/8/2007 12:21:11 PM , Rating: 2
Seagate Barracuda 7200.10 uses four 250GB platters and features 7200rpm motor, 16MB cache as well as Serial ATA-300 interface with native command queuing.

Density=180 Gbits per square inch


By Topweasel on 6/8/2007 12:49:54 PM , Rating: 2
Keep in that they are probably speaking about the single side of a platter as I am sure they have already hit 125GB on a single side. That would mean that they could hit 1TB in a 2 Platter drive.


By namechamps on 6/9/2007 9:11:19 PM , Rating: 2
No they are not.
180Gbit/sq inch = 125GB per side of the platter.
250GB for full plater (double sided).
1TB possible w/ 4 platters.

Higher capacities per platter will lead to lower prices. Expect 500GB drives (based on 2 platters vs 3 currently) to drop down to <$80 price point. Single platters drives now allow 250GB for <$50. Current 750GB Seagate is 4 platters and now it can be built with 3 likely bringing prices down 20% or more. The flagship using 4 platters will allow Seagate to price their 1TB drive cheaper than Hitachi that uses an extra platter for the same capacity.


Bigger isn't always better!
By AlmostExAMD on 6/8/2007 10:13:07 PM , Rating: 2
I want speed, All they do is give us even bigger hard drives.
I mean seriously how many people need even more space to fill with useless crap they hardly look at, Is it 5%, 10% come on WE WANT SPEED.
I and many friends already have enough pr0n to last 3 life times, lol!
Hdd's are still amongst the slowest component of pc's, My games and apps wopuld benefit from a faster affordable desktop drive, Raptor is good but still no where near fast enough.




RE: Bigger isn't always better!
By kensiko on 6/8/2007 10:29:13 PM , Rating: 2
... SSD is THE solution.


RE: Bigger isn't always better!
By RamboZZo on 6/9/2007 9:44:32 AM , Rating: 2
Faster hard drives than the raptor have been around for a long time. You just need to dish out big bucks to go with a scsi 15,000rpm drive. I doubt we'll see faster ATA drives for some time since there is little profit in it, even if the enthusiast community will be all over it.


By PseudoKnight on 6/10/2007 6:46:07 AM , Rating: 2
Um... higher density platters = faster performance, not just more HD space. This is the primary method for increasing performance for these types of hard drives.


Combine this with a 10,000RPM motor!
By vtohthree on 6/8/2007 4:13:47 PM , Rating: 2
You combine this single platter with a 10k RPM motor and at least 32mb of cache, and this bad boy will set new standards across the board!

Hope some company be it, seagate, WD, or Hitachi, is listening!




By mindless1 on 6/9/2007 7:00:12 PM , Rating: 2
Here you show your misunderstanding. You can have higher density and more time to read, or higher RPM. Not both (and remain reliable).


"If you can find a PS3 anywhere in North America that's been on shelves for more than five minutes, I'll give you 1,200 bucks for it." -- SCEA President Jack Tretton









botimage
Copyright 2010 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki