Print 64 comment(s) - last by rippleyaliens.. on Jan 23 at 1:39 PM

Seagate's Savvio 15K hard disk drive

Seagate's Savvio and Cheetah at 15K compared

The Savvios compared at 10K- and 15K-RPM
Seagate claims that its newest 2.5-inch 15K-RPM hard drive is the fastest in the world

Seagate Technology has announced what it claims to be “the world’s fastest hard drive” – the Savvio 15K with a seek time of a mere 2.9 ms. The new 15K-RPM addition to the Savvio family offers a number of advantages over 15K-rpm 3.5-inch drives including size and weight (due to 2.5-inch form factor), 30% decrease in power consumption (5.8 watts at idle), and reliability (1.6 million hour MTBF).

 “Seagate is committed to delivering solutions that will meet the needs of today’s demanding IT environment, and no product demonstrates this better than the Savvio 15K drive,” said Sherman Black, senior vice president and general manager, Seagate Enterprise Compute Business. “The development of the 2.5-inch Enterprise form factor represented a new way of thinking. Now, with the added number of performance and capacity choices offered, many of the leading enterprise system makers are transitioning from 3.5-inch to 2.5-inch form factor enterprise solutions.”

The move to small form factor enterprise disk drives was driven by data center requirements for greater storage performance density while focusing on lowering power consumption and cooling costs.

“The trend in IT is to scale down the physical size of components while scaling up capacity and performance,” said John Rydning, IDC’s research manager for hard disk drives. “Seagate’s first generation 2.5-inch 15K-rpm HDD is fitting with this trend, delivering fast I/O performance in a small package to meet the needs of demanding server applications.”

The Seagate Savvio 15K drives are shipping today in 36GB and 73GB capacities through OEM customers. HP is now shipping Proliant systems with 15K Savvio drives. The Savvio 10K.2 drive will launch in the channel this quarter as a replacement product for Savvio 10K.1 and as a transition path from Cheetah 10K.7 drives.

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

cool... i guess
By NotAok on 1/16/2007 9:14:20 AM , Rating: 2
I'm not too familiar with hard drives, but why are these faster ones so small? (as far as storage capacity) It's great that they're getting faster but it seems to be at a huge cost... And are the speeds even that noticeable?

A month ago I purchased a 500 gig seagate hard drive for $135. To me, that's an amazing deal. I can't see spending more than that on a drive that, while it may be faster, holds so little.

RE: cool... i guess
By daniel1113 on 1/16/2007 9:26:28 AM , Rating: 3
That's because you're looking at these drives from the point of view of a home user, not a corporate user. Yes, the increase in speed is quite noticeable, and a couple dozen of these drives in a server would be quite useful to many organizations.

RE: cool... i guess
By sviola on 1/16/2007 9:27:44 AM , Rating: 2
In enterprise environment which require many I/O operations with huge ammounts of data, you'll see the benefits from faster drives.

Home users don't, usually, require high speed drives as they don't, usually, do intensive disk operations (you'd see better gains improving memory capacity, cpu or gpu). But eventough, if you use a Raptor as your primary disk you'll see some increase in intensive disk operations (like level loading on games, virus searching and defragmentation).

RE: cool... i guess
By Hydrofirex on 1/16/2007 1:06:38 PM , Rating: 2
I wasn't sure about spending 200+ on a 150 Gig Raptor drive, but I do notice a very, very significant difference when loading programs and games. I'm always one of the first 3 people in a level, and I do spend a lot less time loading levels over-all on even non-online games.


RE: cool... i guess
By SteffenThomsen on 1/16/2007 9:31:11 AM , Rating: 3
This drive is not designed with the private user in mind. The selling point of this drive is the incredibly low seek times, physical size, power consumption and MTBF.

The low seek times improve the performance in many server environments because of the many random reads and writes (unlike the more sustained transfer oriented usage pattern of private users).

The physical size allows more drives in less space, which enables more drives in less space which is especially good in blade servers as they're very cramped already.

The power consumption reduces the need for cooling in large scale server farms.

The MTBF improves overall data security.

The amount of available storage is not as important for this kind of drive as the performance aspect. With today's cool technologies like SAS it is easy to incorporate large SATA drives for storage in your servers to take care of the storage needs, and low capacity, high speed drives like the Savvios for performance.

RE: cool... i guess
By AllYourBaseAreBelong2Us on 1/16/2007 10:20:08 AM , Rating: 2
MTBF stands for "Mean Time Between Failure". It is the average time a component works without failure.

RE: cool... i guess
By zsdersw on 1/16/2007 10:56:52 AM , Rating: 3
Data security can also mean something to the effect of: "secure the data's availability". A higher MTBF means the data is more secure in the sense that it's less likely to be lost due to device failure.

RE: cool... i guess
By Mday on 1/16/2007 11:10:35 AM , Rating: 3
I think he meant integrity, versus security.

RE: cool... i guess
By Oregonian2 on 1/16/2007 3:10:24 PM , Rating: 3
Another way to look it is that after the MTBF amount of time, half of one's units will have failed and half still work.

RE: cool... i guess
By JCheng on 1/18/2007 2:57:44 AM , Rating: 2
No, that would be Median Time Between Failure. ;)

RE: cool... i guess
By cgrecu77 on 1/16/2007 9:38:59 AM , Rating: 3
these hdds are used for servers, most servers that use this kind of disk are dedicated to a single website. There are very few websites out there that require more than 30GB of space - and the biggest bottleneck for a wisele configured server is usually the disk system.

for storage you don't need speed, you need capacity ...

RE: cool... i guess
By OrSin on 1/16/2007 9:58:35 AM , Rating: 2
So many people wrong. The one guy got it that these are for servers. Servers are all going to small form factors. Blade 1 and 2 U servers. Even data pass 4-8 processors are using very small forms. Servers are all all going to NAS or SAn or SAS. Its getting harder and harder to get big raid arrayies to support servers. Its going to centralize management. That measn these drive are made to boot servers and unsder some circumsances run small data base at least ont he server end. On the storge end this are very different from just 5 years ago. Before SAN was ll rate by size of storge but thats is no long the case. Size is easy OI processes is not. WE current ha server with 12 TB and it not even 50% full and the server is bottled necked every day. New SANs need IO and the easy wasy to get it more and smaller drives. Also the smaller driver have better seek time. I hope I explained it clear.

RE: cool... i guess
By NotAok on 1/16/2007 10:22:39 AM , Rating: 2
oh i see, thanks guys. I guess I primarily was thinking of a home user rather than business use. I was in that mindset simply because I recall a friend or two going nuts over the fact that their pc had a 10,000 rpm hard drive, yet it barely had any storage.

RE: cool... i guess
By Axbattler on 1/16/2007 7:28:22 PM , Rating: 2
Well, a fast HD can benefit the home users too. Having a fast drive as an OS drive is a plus. It's poor value per GB, but there are people from the enthusiast crowd who would pay a good chunk of extra to maximise performance. Prior the Raptor, a relatively small proportion of enthusiasts would go SCSI to have access to HD faster than IDE solutions. In fact, I wouldn't be surprised if there are still enthusiasts taking that path (the Raptor is very fast, but not the fastest).

And if capacity is needed, there is always the option of getting a large SATA HD for data storage.

RE: cool... i guess
By wrekd on 1/16/2007 11:09:20 AM , Rating: 2
"for storage you don't need speed, you need capacity ... "

I work on an Enterprise SAN and find this statement completely false. If you want to synchronously replicate data offsite, you need speed. All of our drives are 146GB Fibre Channel drives for EMC Symmetrix systems.

This drive featured here would ideally be use for the OS.

RE: cool... i guess
By masher2 on 1/16/2007 11:24:29 AM , Rating: 2
> "for storage you don't need speed, you need capacity ... "

I had to laugh at that also. I have a 5TB scientific dataset, most of which has to be updated daily. Try doing that with a slow array of drives. For transactional systems, the low latency of these drives would be even more important.

RE: cool... i guess
By patentman on 1/16/2007 12:21:09 PM , Rating: 2
Actually the problem with storage capacity and high rotational speed is a technical issue. It has nothing to do with marketing. See my post below.

RE: cool... i guess
By Comdrpopnfresh on 1/16/2007 10:44:18 AM , Rating: 2
its the other way around- they're faster because they're smaller- less of a radius means a motor has to provide less torque to spin the platters at the same speed- faster reads are because the head has to move a lot less of a distance...

RE: cool... i guess
By kextyn on 1/16/2007 11:30:36 AM , Rating: 2
That's not exactly true. The seek time is probably faster because of the shorter distance for the head to travel. But sustained data rates will be slower because the outside edge of a 2.5" platter will be spinning slower than the outside edge of a 3.5" platter if both of their motors are spinning at 15k RPM.

RE: cool... i guess
By mindless1 on 1/16/2007 11:54:55 AM , Rating: 1
They're faster because of the higher RPM. That higher RPM can be maintained more reliably, with higher MTBF because the platters are smaller.

RE: cool... i guess
By Axbattler on 1/16/2007 7:16:38 PM , Rating: 2
If RPM were the sole variable in HD performance, then all 7200RPM drives would perform alike, all 10k RPM drives would perform alike, etc.

But it's nowhere the case. I had a venerable 2nd Gen Cheetah (10k RPM), and while it was one of the fastest HD at the time, a modern SATA drive would stomp all over it in desktop applications.. and most likely server applications as well. Without even going that far back, compare the three generation of Raptors. Or even the 'silent' improvement made to the 2nd gen Raptors.

Spinddle speed is a highly influencial variable. But density, firmware, seektime, cache, number of platters can all affect performance (some more than others).

RE: cool... i guess
By IntelUser2000 on 1/16/2007 8:36:40 PM , Rating: 2
a modern SATA drive would stomp all over it in desktop applications.. and most likely server applications as well.

(FU AT/DT and the stupid quote system)

Probably not. Newer SATA products won't make up for such a big difference. And newer server products based on SAS will probably be faster. If SCSI/SAS products were slower than SATA drives in workstation/server apps as much as it did in PC apps, nobody will buy them.

RE: cool... i guess
By Axbattler on 1/17/2007 9:01:30 PM , Rating: 2
Keep in mind that I am talking about a drive from 1998.

Even the fastest SCSI drive is not future proof ;)

RE: cool... i guess
By rippleyaliens on 1/23/2007 1:39:57 PM , Rating: 2
Well, here is the Real Deal Info. A 15 k drive, vs a 10k raptor in a Desktop PC. It is hare to really show teh clear differenc. BUT if you put the Raptor in a Server environment, in which you have multiple READS /Writes. THat extra 5k, shows a clear difference. More so, the disk I/O, are just clearly supperior with a faster hard drive. MEaning, doing more than just copy / pasting.
Disadvantage with SATA is that for each drive, there needs to be a channel. VERSUS with SAS, each channel can support multiple drives.

There are some good SAS/SATA controller out there. The difference comes also, when you RAID the drives. I have at work, a 12 drive 500gb, sata array. Using the SEAGATE lastest GEN, with the command q, etc... Now in a raid 10 array, i get 3tb, BUT versus my 6 drive SAS 15k raid 10 array, the SAS litteraly SMOKES my sata array. Sata was not designed for anytype of HEAVY sustained READ / WRITE, condition. Now some PC stuff, but heavy workstation / server class IO.
With VISTA now, they are trying everytihgn in the world to speed up the OS, but it is still bottlenecked by the DISK speed / I/O. You best believe i will get these new drives. IN a 2 u CHASIS, i can put 12 drives. IN a raid 10, that is a sustained 900 (non cached)I/O Per second. 4-500 MB sustained Transfer rate. yah, that is very fast.

RE: cool... i guess
By patentman on 1/16/2007 12:14:02 PM , Rating: 2
"I'm not too familiar with hard drives, but why are these faster ones so small? (as far as storage capacity)"

See my posts on the anandtech forum (under klaviernista) at:


They should answer your questions.

RE: cool... i guess
By Madzombie on 1/16/2007 3:37:14 PM , Rating: 2
I always wondered why they never pushed speeds above 15,000RPM. I assume all 15K drives have 2.5" platters. If this is the case then faster drives might be possible with smaller platters, say 20,000RPM drives with 1.8" platters. Capacity would be reduced further but with even smaller platters the seek times should be proportionally shorter.

RE: cool... i guess
By Missing Ghost on 1/16/2007 7:04:06 PM , Rating: 2
Now that they use 1.8" platters in server drives, I suppose such a 20krpm product will soon be released.

RE: cool... i guess
By patentman on 1/17/2007 7:43:09 AM , Rating: 2
US Patent 6819531 to Shiroishi teaches a hard disk apparatus having a 50 MB/s transfer rate where the platter is spun at 25,000RPM. The recording density is fairly low (5GB/square inch) compared to slower drives (50-100gb/siquare). But it is fast!

Higher rotational speeds are definitely possible, but eventually new platter substrates will have to be developed in order to handle the g's

sas controller
By Lazarus Dark on 1/16/2007 9:51:20 AM , Rating: 2
Just wish sas controllers were cheaper, maybe even standard on enthusiast mobo's if the cost was the same. I wouldn't use many sas drives but it would be nice to have one or two of these drives for os and games.

Then again, why won't wd make a 15k raptor. Is there some limitation with sata that prevents 15k drives on sata?

RE: sas controller
By phil126 on 1/16/2007 10:01:09 AM , Rating: 2
a SAS drive can work on a SATA and vice versa. There are certain features that SAS has that SATA does not have.

RE: sas controller
By GaryJohnson on 1/16/2007 1:54:37 PM , Rating: 2
SAS drives will not operate on a SATA controller and are keyed to prevent any chance of plugging them in incorrectly

RE: sas controller
By IntelUser2000 on 1/16/2007 10:18:48 AM , Rating: 2
Just wish sas controllers were cheaper, maybe even standard on enthusiast mobo's if the cost was the same. I wouldn't use many sas drives but it would be nice to have one or two of these drives for os and games.

Then again, why won't wd make a 15k raptor. Is there some limitation with sata that prevents 15k drives on sata?

But SAS is not better than SATA for PC environments. Performance differences prove it:

The U320 drives suck until the database benchmark which then even the Raptor looks like a drive from the 1980s.

This is just like people complaining lack of FB-DIMM+5000X performance, which is bullshit as 5000X is made to perform in server environments.

RE: sas controller
By IntelUser2000 on 1/16/2007 10:23:18 AM , Rating: 2

Again, Put the Wildcat graphics card that's kicking PC graphics card's ass to a video game and it'll suck throughly.

(Anandtech/Dailytech must use the most stupidest programmers for their web designs. Why can't they have no window popping up everytime you want to quote something?? Plus, can't they put a fcking edit button??)

RE: sas controller
By Ard on 1/16/2007 11:32:01 AM , Rating: 2
Except that 3DLabs has all but stood still with respect to the Wildcat line (they've pretty much thrown in the towel in the workstation market), whereas NVIDIA has continued to develop the Quadro (and I'd imagine that a G70 based, and definitely a G80 based, Quadro could best the Realizm 800).

RE: sas controller
By IntelUser2000 on 1/16/2007 8:23:02 PM , Rating: 2
Except that 3DLabs has all but stood still with respect to the Wildcat line (they've pretty much thrown in the towel in the workstation market), whereas NVIDIA has continued to develop the Quadro (and I'd imagine that a G70 based, and definitely a G80 based, Quadro could best the Realizm 800).

I am assuming you are smart enough to know the point I was trying to get across. A server product is gonna perform better in the intended market than the other ones, like PC.

RE: sas controller
By Lazarus Dark on 1/21/2007 2:00:17 PM , Rating: 2
But SAS is not better than SATA for PC environments.

What I meant was I would like a sas for the 15k speed, not for sas features.

By KingstonU on 1/16/2007 10:25:00 AM , Rating: 1
Companies are going to stop bothering with making faster traditional disk drives now that Solid State Drives (SSDs) are just around the corner. SSDs are the same price as this 15K rpm disk drive or possible slightly higher, but prices will drop quickly and capacities will increase quickly while they are many times faster and use a fraction of the power consuption.

Anandtech and Dailytech have covered this:

RE: Pointless
By Master Kenobi on 1/16/2007 11:06:56 AM , Rating: 2
Current problem with SSD is the write speeds. They write very slowly compared to server class drives. Seek and Random Access time is great, but not so great on the Write time, and with lots of writing done to servers every second, SSD's can't keep up right now. Servers are going to stay magnetic for the time being.

RE: Pointless
By wrekd on 1/16/2007 11:35:18 AM , Rating: 2
SSDs will not even have 50% market share in 7-10 years

Faster traditional drives are and will continue to be produced for a long time. Cache and interface speeds definitely will increase and Hybrid HDDs will be the norm. The SATA specification already increased from 1.5Gbits to 3.0 Gbits and 6.0 is coming.

This is a great advancement as it allows more drives to fit in a smaller space while using less power and does not impact performance.

RE: Pointless
By drank12quartsstrohsbeer on 1/16/2007 1:10:40 PM , Rating: 2
SSDs will not even have 50% market share in 7-10 years

I'm not so sure about that. Other than increasing storage capacity, traditional hard drives have stagnated.

The only reall selling point is the cost per GB of the new drives. But if a user only needs 200 GB, why buy a terabyte size drive?

The unit cost to manufacture a hard drive isn't going to get a whole lot cheaper in the future. There aren't any more economies of scale to take advantage of, and there is no way to reduce the cost of the raw materals and mechanical components a HD requires.

Flash memory, on the other hand is much less technologically mature, new flash technologies and manufacturing methods are signifigantly improving speed and capacity.

Us nerds always forget that the typical computer isn't pushed that hard. Most computers sold have Windows, Office and internet explorer installed, and thats all they EVER have installed. 100GB will be fine for this type of computer, for its entire operating life. In a couple years 100 gb flash drives will give comparable performance and price to platter based drives. As soon as that happens, its all over.

RE: Pointless
By FITCamaro on 1/16/2007 1:48:52 PM , Rating: 2
You can make the interface as fast as you want. Its not the interface thats slowly down regular hard drives. Its the fact that the platter only spins so fast. ATA100 is more than enough bandwidth for any regular hard drive. The only thing SATA has given us is thinner cables, NCQ, and a few other minor improvements to the efficiency of the drive getting data off the disc. The only thing that will make traditional hard drives faster is to pack more data in less space.

RE: Pointless
By Motley on 1/16/2007 3:09:38 PM , Rating: 2
If you only look at sustained throughput then you are correct, but you are forgetting that the drives also have a decent amount of cache on them that a slower (shared) ATA100 bus is slowing down considerably. If you actually use both connections on the ATA100 interface, then you'll run out of bandwidth even on sustained throughput (some drives can do a sustained throughput of over 80Mb/sec), and they keep getting faster.

RE: Pointless
By Oregonian2 on 1/16/2007 3:12:44 PM , Rating: 2
The new vertical magnetic recording that Seagate and others are now using increases the data density quite a bit on the platters and that runs up the sustainable data rate (until a seek is required anyway).

By Axbattler on 1/16/2007 9:11:07 AM , Rating: 2
I wonder how they came up with the 11% fewer drive failure. Compared to what? The average? The(ir) 'next most reliable drive'? While I appreciate any claims for improved reliability, I am going to have to take it with a pinch of salt.

Other than that, it looks good. Still, drives of this capacity will probably have to compete with SSD in due time.

RE: Reliability
By Flunk on 1/16/2007 9:24:45 AM , Rating: 2
It says 15% fewer failures than the Seagate Cheatah 15K.5 on the charts beside the text.

RE: Reliability
By Axbattler on 1/16/2007 9:37:30 AM , Rating: 2
But claiming that it's a 'world record' implies that it is the most reliable drive doesn't it?

RE: Reliability
By HammerZ on 1/18/2007 6:42:23 PM , Rating: 2
It's based on MTBF numbers of the components used on the unit. Seagate probably was able to reduce the number of components or use components w/ higher MTBF numbers. MTBF is a theoretical number, which may or may not be the same as field failures.

Nice and fast but...
By cyberguyz on 1/16/07, Rating: 0
RE: Nice and fast but...
By Goty on 1/16/2007 10:08:22 AM , Rating: 2
You have to remember that these are operating at much higher rotation speeds, so fewer platters must be used to reduce rotational inertia. Most drives operations at 10-15K RPM use only one platter, so this kind of data density is just fine.

RE: Nice and fast but...
By patentman on 1/16/2007 12:22:38 PM , Rating: 2
one word..microwaviness. See the links in my post above.

RE: Nice and fast but...
By jordanclock on 1/16/2007 7:10:41 PM , Rating: 2
These aren't for storage. Capacity isn't as important as speed and reliability for the role they fill. Those consumer class drives probably fail quite a bit more often than these drives. Look at the capacities of the line they are replacing, the Cheetah. Those have 73GB and 146GB capacities. I wouldn't be surprised if the next generation of these have similar capacities, but at the same time, the demand may not be there for such sizes.

RE: Nice and fast but...
By Axbattler on 1/16/2007 7:55:37 PM , Rating: 2
The Cheetah 15k.5 comes in 300GB flavour too. And I suspect that there is still a place in the market (e.g. video editing) for a moderately large HD, with insane transfer speed.. Which is exactly what the Cheetah 15k.5 is. It doesn't break any record for random access and I/O (though it is far from being a slowpoke), but the 82.6MB/sec to 135MB/sec transfer (sans RAID-0) puts it on a fairly confortable position right now.

Still. It -is- interesting to see that they are now branding the Savvio, previously a 10k RPM drive as their new benchmark setting drive.

2.5" right?
By Comdrpopnfresh on 1/16/2007 10:48:13 AM , Rating: 2
if these are already in the 2.5" form factor, why not put them in notebooks? faster reads, better shock protection, less power consumption, marginal capacities- they could give ssd a run for its money while its in its infancy (b/c of slow r/w and capacity, not to mention price)

RE: 2.5" right?
By Anonymous Freak on 1/16/2007 12:36:06 PM , Rating: 2
It's less power consumption than desktop 15k RPM drives. Still noticeably more than notebook 7200 RPM drives. Again, all the benefits are vs. previous server-class drives, *NOT* vs. current notebook drives.

Not to mention the fact that I haven't seen a single notebook with an internal SAS controller. (Serial Attached SCSI, the SCSI equivalent of SATA.) And while SATA drives are compatible with SAS controllers, the reverse is not true. A SAS drive will only work on a SAS controller.

RE: 2.5" right?
By FITCamaro on 1/16/2007 1:45:29 PM , Rating: 1
Also the heat generated by such as fast drive in a notebook would be insane. 7200 rpm drives are already questionable in notebooks. 15k rpm drives in a notebook would probably melt the case.

RE: 2.5" right?
By thecoolnessrune on 1/16/2007 11:19:11 PM , Rating: 2
Power consumption at idle of most notebook drives are .5 watts. So your still talking about 11x more power to sit there and idle. And at 15k RPMS, you'd drain the battery in a heartbeat. When it comes to notebook batteries, even half a watt makes a world of difference in run time.

A major improvement of a huge bottleneck.
By TheBluePill on 1/16/2007 9:42:44 AM , Rating: 3
A major improvement of a huge bottleneck.

Hard drives have been a huge bottleneck for all PCs since the get-go. Going from a 5.4K to 7.2K RPM Drive is a major boost, even for home users. The high priced 10K raptors and the like are another “trim” off the bottleneck. At 15K, things are even better.
Everyone should be happy to gain even a little bit of disk performance. I think this is just as important for the home user as the business user as it makes the PC experience and usability that much better. Even if you are only cutting 1-2 Seconds off a disc read when opening a large word document or a video on your hard drive is worth it. It’s a step in the right direction and the cost will come down as the technology and demand matures.

On the flip side, I do see flash based drive demand outpacing disk based storage over the next 5 years. You have much better power consumption, much faster access time and much lower failure rate. But these 15K drives are a great stop-gap.

RE: A major improvement of a huge bottleneck.
By Pirks on 1/16/2007 3:48:05 PM , Rating: 1
look, just put two 16GB SSDs in a raid0 stripe, you get double the write speed (so this bottleneck's gone) and you can say bye bye to this 15k crap. ok, if you're uber 1337 and have dough - get FOUR such 16GB or even 32GB SSDs, stripe 'em... and see 15k seagates waaaaay behind ya, desperately trying to catch up :)))

that's right now. don't even wanna know what second gen cheaper SSDs will bring us. I gonna save some money, stripe a couple of those and forget about Vista and games taking forever to load. SSD, here I come baby. seagate, kiss my middle finger :P

By Hawkido on 1/19/2007 4:40:34 PM , Rating: 2
Yeah I thought about that. However, I also thought about the fact that while newer PC's can boot off of USB. It cannot setup software RAID on USB before it boots, so therefore it cannot boot to a USB stripe without a hardware RAID USB controller.

By Souka on 1/16/2007 10:11:18 AM , Rating: 2
so can these drives go into a laptop computer?

I know a few developers / graphics folk that would really benifit from such a drive. They use DTR laptops (the big 17"+ models that weigh a ton).

RE: ???
By Hulk on 1/16/2007 10:26:52 AM , Rating: 2
Assuming a 3.5" and 2.5" platter, at the outside edge of the platter the velocity of the platter is only about 7% faster than a 3.5" 10k drive. Of course the speed advantage would increase as move toward the spindle from there and the access time of the 15k is theoretically 50% faster than 10K.

By GhandiInstinct on 1/16/2007 5:09:01 PM , Rating: 2
Is this SATA?

By jordanclock on 1/16/2007 7:03:38 PM , Rating: 2
No, they're SAS (Serial Attached SCSI). The connector for some SAS drives is the same, but most are different, I believe. Even if the connector is the same, though, the electrical and digital workings are different. Basically... Don't expect to see this in a bleeding edge gaming rig for a while.

By abcslayer on 1/16/2007 11:16:44 PM , Rating: 2
I love to see the high MTBF and low unrecoverable sector error rate. I wonder how manufacturer deal with low write cycle of flash-based drive (SSD)? (are they keep other 64GB flash onsite to replace for dead flash cell?

"Intel is investing heavily (think gazillions of dollars and bazillions of engineering man hours) in resources to create an Intel host controllers spec in order to speed time to market of the USB 3.0 technology." -- Intel blogger Nick Knupffer
Related Articles

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki