Western Digital Working On a 20,000 RPM Drive

MrKaos writes "Western Digital seems to be preparing for the onslaught of solid-state drives set to impact its market by developing a 20,000 rpm hard drive. Similar to the VelociRaptor line of drives, the new drives are speculated to be offering lower capacity as a tradeoff for faster seek and write times." This report out of Taipei is the only word on the rumored WD 20K drive. It's said to be a 2.5" drive in a 3.5" enclosure, for efficiency of cooling — the arrangement the Register enjoyed poking fun at when the 10K drive was upgraded last month.

Comments

Seagate responds

[Leuf]

We've taken the next step by mounting our 15,000 rpm drives in an external enclosure which then spins the drive at a further 10,000 rpms, for a total system speed of 25,000 rpms. Initial benchmarks are very promising!

Re:Seagate responds

[sunami]

We've taken the next step by mounting our 15,000 rpm drives in an external enclosure which then spins the drive at a further 10,000 rpms, for a total system speed of 25,000 rpms. Initial benchmarks are very promising!
Pretty sure this is a general relativity question, so it would be less than 25,000. Come on get your science right.

Re:Seagate responds

[Firehed]

Before now, nobody understood why I have all of my computers sitting on top of turntables. Now I'll just point them to your post, since they couldn't fathom what I meant when I said it makes it run faster.

I had also tried mounting them in a paint can shaker to get at least another 15Hz out of the CPU, but I couldn't stand the noise.

Re:Seagate responds

[EdIII]

Come on get your science right.

You don't get +5 Funny for "getting your science right".

Re:Seagate responds

[davester666]

I find putting my computers on a treadmill makes them run even faster than on a turntable. As a bonus, treadmills are much sturdier than turntables, so they last much longer before they need repair.

I also have stopped getting dizzy trying to look at the pulsing light on my MacBook Pro to see if it's sleeping or not since switching to a treadmill.

Re:Seagate responds

[Enderandrew]

http://funroll-loops.info/

Paint flames on it, install Gentoo, and use --omg-optimized

immovable object?

[seeker_1us]

I wonder if these really fast hard disks will have to be kept stationary. More specifically: I wonder if conservation of angular momentum (manifested, for example, in gyroscopic precession) becomes a real issue if any torques were put on a spinning disk.

Re:immovable object?

[pushing-robot]

Yes, well, giant fighting robots need enterprise-grade hardware too.

Re:immovable object?

[TechForensics]

The smaller diameter / mass will tend to reduce bad effects from conservation of angular momentum.

Re:immovable object?

[Z00L00K]

Even though they are intended to be used in server hardware where they are going to be kept stationary you will also be able to find users that are going to use them in their home computers or in servers that are on the move.

This means that the gyro effects are worth to consider. Also considering my experience from WD disks I'm not sure that I would want to use them for anything reliable.

For a solution where speed is important but the data itself can be re-created or of less critical value they can be OK.

Re:immovable object?

[audunr]

Once you hit 20K RPM, the platters stay put while the universe is spinning.

Solid State

[c0d3r]

I'm wondering why they are still going in this direction, as hard drives are the slowest part of a computer. Why hasn't a solid state / flash ram approach taken over? Is it feasible to have a hybrid solid state/mechanical solution?

Re:Solid State

[jay-be-em]

Economics.

Re:Solid State

[c0d3r]

Apparently http://en.wikipedia.org/wiki/Solid-state_drive answers my questions:

Price - as of mid-2008, flash memory prices are still considerably more costly per gigabyte than are comparable conventional hard drives: around USD 3.50 per GB[10] compared to typically less than USD 0.40 for mechanical drives.[11]

Capacity - although currently far lower than that of conventional hard drives, SSD capacity is predicted to increase rapidly, with experimental drives of up to 1 TB in test.[12][13]

Higher vulnerability to certain types of effects, including abrupt power loss (especially DRAM based SSDs), magnetic fields and electric/static charges, in comparison to normal HDDs (which store the data inside a Faraday cage).

Limited write cycles - flash-memory cells will often wear out after 10,000-100,000 write cycles[citation needed], while high endurance cells may have an endurance of 1-5 million write cycles (many log files, file allocation tables, and other commonly used parts of the file system exceed this over the lifetime of a computer.[14] Special file systems or firmware designs can mitigate this problem by spreading writes over the entire device (so-called wear levelling), rather than rewriting files in place.[15] Today's drives can last up to 20 years with average usage.[dubious - discuss] An example for the lifetime of SSD is explained in detail in this wiki.[dubious - discuss] SSDs based on DRAM, however, do not suffer from this problem.

Slower write speeds - as erase blocks on flash-based SSDs generally are quite large, they are far slower than conventional disks for random writes and therefore vulnerable to write fragmentation,[16] and in some cases for sequential writes.[6] SSDs based on DRAM do not suffer from this problem.
Lower storage density - hard disks can store more data per unit volume than DRAM or flash SSDs, except for very low capacity/small devices.

Higher power consumption at idle or under low workloads laptop battery runtimes decrease when using an SSD over a 7200 RPM 2.5" laptop hard drive,[17] flash drives also take more power per gigabyte.
RAM based SSD require more power than hard disks, both operating and when turned off.[18]

Re:Solid State

[rootofevil]

that bit about power consumption was partially recanted by Toms.

apparently they didnt do their homework well enough.

color me surprised.

not to mention the article basically says that current drives have almost no power saving features and performance was on par to resulting in slightly more consumption, whereas platter drives have had decades to develop power saving features.

i expect this 'result' to be completely wrong in the next couple product cycles. the intel 'mass market' drives already advertise significant power saving potential and are the first consumer component offered.

Re:Solid State

[omfgnosis]

It seems the long-term answer would be DRAM-based, with a battery to help mitigate corruption on power loss. Obviously older DRAM technology hasn't gone down in price/GB, but presumably that's due in part to production going down. I'd love a boot disk using old 100 MHz DRAM chips, if it could be made dependable and as affordable as flash drives.

Re:Solid State

[eelke_klein]

Maybe because western digital is a harddrive manufacturer, not a flash memory manufacturer. They do not have the know how to make flash chips they could buy them but it would be hard to compete against the manufacturers that bake their own flash memory.

Re:Solid State

[maynard]

If you want to go RAMDISK, you're better off buying an 8 or 16 DIMM slot motherboard and installing 32GB or 64GB of physical RAM on the host itself. The the main system bus is far faster than pcie. Use LVM to snapshot and then backup your RAMdisks to physical disk at regular intervals. Put /usr, your SQL DBs, and any other dynamic data in RAM and enjoy just silly performance. Works great in combination with Xen.

And don't forget your dual redundant power supplies and a good working UPS.

Is there a point to this?

[Ostsol]

Is there still really a point to huge RPMs? As data density increases, speed should increase naturally. Move over the same distance at the same speed on a drive with twice the density should mean that one has read twice the data in the same amount of time -- therefore reading speed is twice as fast, right? This should even work on low-capacity drives by simply using small, high-data-density disks.

Re:Is there a point to this?

[hedwards]

No, not right, that's assuming that the writes are being done sequentially. Hard disks are random access devices, and while they can definitely do sequential reads and writes, and quite a bit faster, as soon as the files are not next to each other or are fragmented you're going to lose that advantage.

Re:Is there a point to this?

[Jeff DeMaagd]

The higher speed drives aren't so much for their sequential transfer rates by themselves, but their random seek rates. They are trying to get high I/O per second rates (IOPS), which is what a lot of servers need to be at their peak.

Re:Is there a point to this?

[Anonymous Coward]

No matter how dense the data is, random access speeds are dominated by how long it takes to move the head to the data and how long it takes to wait for the data to rotate under the head. A smaller platter will mean it takes less time to move the head on average, but the only way to get the data under the head faster is to increase rotational speed.

A 7200 rpm drive has an average 8.3ms rotational latency; a 15k rpm drive is 4ms, and a 20k rpm drive is 3ms. In other words, this speed increase could enable the drive to do 10% more random I/Os per second.

dom

Re:Is there a point to this?

[gardyloo]

In other words, this speed increase could enable the drive to do 10% more random I/Os per second.

We at the NSA are interested in things which are more random, and would like to subscribe to your newsletter.

Re:Is there a point to this?

[Detritus]

You can't just wave a magic wand and double the density. Many of these things are interrelated in a complex manner. Increased density requires new head designs, new and improved electronics, new coatings for the platters, etc.

Re:Is there a point to this?

[Anonymous Coward]

The higher speed drives aren't so much for their sequential transfer rates by themselves, but their random seek rates.

The secret goal is to use the drive as a flywheel, and power a laptop.

Re:Is there a point to this?

[FromFrom]

Those are the maximum rotational latencies, not average. On average you will wait half a rotation and thus making the average rotational latency of a 15K rpm drive 2 ms.

Re:Is there a point to this?

[MeepMeep]

The problem is that going from 15k drives to 20k drives means going from roughly 150 IOPS to 200 IOPS.

That's it.

SSDs are getting 20,000 IOPS, and some specialized SSD and RAM systems like FusionIO and Violin memory are getting 120,000 to 1,000,000 IOPS (these are not typos)

Sure, SSDs are expensive but spinning disks are beginning to look more and more like an evolutionary dead end, at least for IOPS.

Add heads?

[macemoneta]

It seems strange to continuously up the rotation speed, adding noise, vibration, heat and shortening the life of the drive. Why not just add another set of heads on the opposite side of the drive? You get many of the same benefits - increased sustained transfer rate, but also reduce the seek and latency. To maintain the form factor, reduce the size of the platters (use 2.5" drive platters in a 3.5" drive).

Re:Add heads?

[vikstar]

Good idea. I for one would prefer to go solid state.

Re:Add heads?

[BronsCon]

I had this same idea, actually, only I thought to have 4 sets of heads, rather than just two.

I also thought of arranging what would essentially be two 2.5" disks in a 3.5" enclosure. These could either act as a stripe for faster, higher capacity data storage, or as mirrors of each other, providing redundancy at the cost of speed and capacity. If the drives in your RAID stripe are mirroring themselves, you needn't worry about mirroring your RAID stripe, no?

Re:Add heads?

[Gazzonyx]

Well, that would add to the number of components that could fail, and require a high speed bus between the two controllers, as well as a shared cache and all the headaches that would bring with it (think SMP caches being ping-ponged). Then you've got to sync your interface to the system bus as well as the new internal buses. On the other hand, you can just crank the knob up to 11 and go 20K RPMs on known, tried and true, technology.

Re:Add heads?

[ka9dgx]

I'd go even further... use the old 5 1/4 "half height" form factor, stack 8 platters in it, with 4 sets of heads, spin it at 5400 rpm to keep the power requirements down to reasonable.
This would give you 8 platters * 2 sides * 400 Gbit/in^2 * 50 in^2 (estimated working area surface area per platter) ==> 40 Terabytes in a single package, with an average access time on the order of 5 millisecond, and a sustainable transfer rate of at least 300 Megabytes/Second.
Even without the 4 sets of heads, that would still be a 40 Terabyte drive!

As far as RAID goes, it's just one drive, it's all or nothing, so don't think it would count as it's own mirror.

--Mike--

Re:Add heads?

[Detritus]

CDC's hard drive division did this years ago. They were later acquired by Seagate. The drives were too expensive for the improvements in performance and were discontinued. It isn't only a set of heads, it's another positioner assembly and a large amount of duplicated electronics. That's more power, heat, PCB space.

Re:Add heads?

[frieko]

Modern hard drives can only read from one head at a time. The tracks are packed in such that thanks to uneven thermal expansion, only one track will be lined up under a head at any given time. But two sets of heads might work as gp suggested.

Re:Add heads?

[ka9dgx]

Each set of heads would have its own servo, so the small variations in distance due to expansion of the case wouldn't matter. I agree that reading from one head at a time is a limitation that has to be lived with, and always will be.

Re:Add heads?

[YesIAmAScript]

Connor actually did this right around the time 3.5" drives started.

http://en.wikipedia.org/wiki/Conner_Peripherals#Performance_issues_and_the_.22Chinook.22_dual-actuator_drive

It could read from either set of heads, but I believe could only write from one set. Writes can be posted in a write-behind buffer, so this didn't impact performance.

Re:Add heads?

[ndevice]

It's been done before, iirc, but they tend to be more expensive, and the multiple heads run the risk of creating unintended harmonics. Most of the time it would be cheaper and faster to use two drives with one set of heads, than one drive with two sets of heads.

Western Digital? Oh good!

[ThatsNotFunny]

Now it can lose my data twice as fast the last one I bought.

Re:Western Digital? Oh good!

[Gazzonyx]

Yeah, it's the last time I'm going to buy one of those things... until the next sale... or if I need to pick one up at Wal*Mart at 3AM. I've been saying that for three years now. Somehow my drives die just in time for the newest WD's to be on sale.

Re:Western Digital? Oh good!

[bigstrat2003]

Bad luck? I've never had a problem with WD, I swear by 'em. One of us is having unusual luck, and I'd prefer to think it's you. ;)

Maxtor, on the other hand... I lost count of how damned many Maxtor drives I've seen die. Single most failure-prone drive manufacturer I've come across. Everyone else, I see a dead drive here and there, nothing serious, but Maxtor is obscene.

We at Gnome Rotary Disks have a better solution

[iminplaya]

The disk is stationary and we spin the case for better cooling.

More Parallelism

[Doc Ruby]

the new drives are speculated to be offering lower capacity as a tradeoff for faster seek and write times

How about if they make drives with very thin platters, but stack them up into individually addressable bit slices of the bytes they store? Then the time to read a single bit from the rotating media could read an entire byte, reassembled in the logic.

Or if the platters can't be that thin, how about sacrificing some storage capacity for say 2x2 platters, which could give 4x parallelism.

That parallel access might stave off competition from solid state drives for a couple extra years.

Re:More Parallelism

[TubeSteak]

How about if they make drives with very thin platters,

http://www.theonion.com/content/node/33930

Make the blades so thin they're invisible. Put some on the handle. I don't care if they have to cram the fifth blade in perpendicular to the other four, just do it!

I was going to do a full rewrite of the article, but you can do it yourself by mentally substituting "hard drive" or "platter" for "blade," "razor," or "shave"

Kenwood TrueX...?

[RudeIota]

Remember those? It kind of reminds me of the multi-head idea you have. Perhaps one of the differences (I think) is that the actual head assembly moved too, to compensate for the disc itself not being able to rotate faster (discs have the potential to shatter above the usual max speeds on current optical drives). I remember seeing a drive rated at 72X back in early 2000.. maybe even 99.. I don't feel like digging up a link though.

They were fast and quiet, but I don't think they make them anymore. I remember the reviews being favorable too, but they were a bit expensive (not outrageously so, IMO).

Re:Kenwood TrueX...?

[karnal]

From what I remember reading, they actually split the laser into several beams that could be read by a single assembly - that way if you were reading one long continuous file, it could effectively increase your read speed since it was reading 3 or more chunks at once.

Random access wouldn't benefit from this as much; for maximum speed you would need to be reading something that spanned the length of all 3 tracks (or more) at once. Writes would be even more interesting in this scenario.

The product will be dead on arrival

[bogaboga]

I am afraid the 20,000 rpm drive might be dead on arrival! Isn't the world "going SSD," whose advantages include faster start-up times, low read latency times, "mechanical" reliability and absolute silence while working?

Laptops have SSDs, next will be desktops increasing chances that Western Digital's product will be dead on arrival.

Re:The product will be dead on arrival

[wisty]

SSD has won most of the phone and iPod tier. Ultra-portables will go next. Then laptops, appliance desktops, gamer desktops, then finally business desktops. This might take 5 years. Only then will servers (which is the market for 20k disks) start to go SSD. 5 years is a long time.

Google might find a use for SSD (using caching and stuff), but your average business running an Oracle database to keep track of their business information wants read-write (not read) performance, which is what these things can do.

Re:The product will be dead on arrival

[bigstrat2003]

When SSDs have large capacities (at least 500 GB) at a comparable price per GB, then I'll consider putting one in my computer. Not before.

Fame

[Tablizer]

Who will be famous as the first consumer to die in a hard-drive failure?

linear speed

[Anonymous Coward]

A 2.5" platter at 20,000 RPM has a linear speed at its outer edge of 257 km/hr (160 miles/hr). Yeah, that's moving along at a pretty nice clip.

Re:Only 20K?

[dfsmith]

The moment you can show me an F1 engine that runs at 20kRPM for 5 years and costs less than $500 I'll get back to you.

Re:They should work on a 20,001 RPM drive

[MostAwesomeDude]

20001: A Speed Odyssey.

Sorry, couldn't resist.

The real future

[caliburngreywolf]

HP has the patent on the memsistor...which is set to revolutionize memory/storage in a big way. Imagine packing a computer away for 2 years, then turning it on, and in half a second, the youtube video you had buffered begins playing. Hard drives will always need to spin up.