Miniature 5400 and 7200 RPM HDDs Reviewed

PReDiToR writes "At Tom's Hardware I found this favourable review of some remarkable Hard Drives. The article points out that with 40GB units suitable for server or desktop use, life with 2.5" drives could be just around the corner. Heat noise and power consumption are all apparently within acceptable tolerances."

Comments

Is that a hard drive in your pocket...

[mikeophile]

Or are you just happy to see me?

Re:Is that a hard drive in your pocket...

[dcw3]

Or are you just happy to see me?

Better than than a floppy

Re:Is that a hard drive in your pocket...

[Dogtanian]

Is that a hard drive in your pocket...
Or are you just happy to see me?

Uh, huh. All this talk of 2.5" disks is going to impress the girls even less than the normal 3.5" (*).

Personally, I have 8", and that's just when it's floppy.

(*) Although posting regularly to Slashdot won't impress them much either.

Re:Is that a hard drive in your pocket...

[evilviper]

Can't it be both?

USB Key's

[Heartz]

With USB keys being much smaller and storage sizes increasing exponentionally,will the spinning disk still be a relavant tool for easy to carry around storage.

USB keys are not only lighter, but you don't even have to worry about it fsckign because you shook it too much while you were on the bus.And they look waaay cooler too.

Re:USB Key's

[tomstdenis]

Yeah, right until after you re-write a sector that 100k'th time......

Normal IDE disks are rated for a billion re-writes at least. [provided the motor lasts that long]. That's why "them there funny rotating disc like objects" are used to store data.

Until they invent a lower-power [recall flash requires around 10V or higher to write, from a 3V source that's a loss!] and longer-lasting high density flash you won't see "them funny discs" replaced any time soon.

Tom

Re:USB Key's

[fr0z]

But USB keys don't have enough space to store my pr0n...

Re:USB Key's

[SubjunctiveSam]

I tried searching for your sig on google, but it returned no matches. It's a cool quote, mind telling me when the National Security Agency said that?

Re:USB Key's

[Alien Being]

"will the spinning disk still be a relavant tool for easy to carry around storage."

It's about 100x cheaper than solid state and storage "requirements" keep going up, so I'd say yes. Disk based digital video recorders will probably catch on at some point.

Re:USB Key's

[zarniwhoop]

too slow - at the mo !

Re:USB Key's

[Phroggy]

Find me a USB key that holds 40GB for under $200 and we'll talk.

Re:USB Key's

[Ed Avis]

Who's to say that USB keys will not be made containing a small hard disk?

Re:USB Key's

[Rolo Tomasi]

Your points are invalid. First, where did you get the idea that flash uses more power than a hard disk? Probably not even if you compare the power consumption on a byte per byte basis. I think it's plain wrong, and the burden of proof for such an outrageous statement is on you. As to your second point, I quote:

NAND - Flash based media uses a complex low level sector virtualization called "wear leveling algorithm" to distribute the wear evenly across the memory array and maximize the number of write cycles it can sustain.

What this means is, your flash gizmo will beign to fail only once all sectors have been worn out. Even with heavy usage, this could take a long time, probably a decade or more, e.g. for a typical CompactFlash card. I had a 4MB CF card that I used on a daily basis for about 5 years in my Psion handheld, and I had no problems with it until I sold the Psion last year. Another factor is, the bigger the flash device, the longer it will take to wear it out. 4GB CF cards are already on the market.

The actual point is, when carrying stuff around, there's a very high probability that it will experience some sort of impact, and you probably know what happens when you drop your hard drive. OTOH, there's no real replacement yet for HDDs in your vanilla PC or laptop. Continuous writing, i.e. having a swap file on flash memory, would probably really wear it out pretty quickly.

Re:USB Key's

[tomstdenis]

I didn't say flash takes more power. I said it takes enough power to make it non-trivial. [e.g. conversion to 10V or whatever it uses].

Second, where do you think these "re-map" sectors come from? They're not free.

Third, reading flash doesn't really wear it down. How much did you write to it? Also I was trying to point out that replacing a HARD DISK with flash isn't entirely viable. I mean what of the logs, swaps and other temp files routinely created [and deleted, and re-created and so on].

Tom

Re:USB Key's

[whovian]

Good point. Don't see why not .. somehow ... eventually, considering that micro-drives are available for use with electronic equipment, notably digital cameras. It looks at though 1 GB is available now, with 4 GB perhaps this fall.

Re:USB Key's

[dsb3]

Find me a USB key that holds 40GB for under $200 and we'll talk.

How big is your keyring? :-)

Re:USB Key's

[der_joachim]

Find me a USB key that holds 40GB for under $200 and we'll talk.

At the moment, the development in capacity for USB drives is way faster than for HDDs. USB storage is getting bigger and cheaper at a fast rate. Furthermore, as another poster already stated, life expectancy isn't that bad either. I would not be surprised if in a few years the good ol' HDD would be replaced by some kind of flash memory.

My computer would be a lot more silent, would probably consume less electricity and it would be lighter (nice for LAN-parties).

der Joachim

Re:USB Key's

[Anonymous Coward]

News from the mysterious future!

Tom St. Denis replaces "them funny discs" with both CANNED MANHAM and BOTTLED MANGOO. More news as events warrant!

Re:USB Key's

[Anonymous Coward]

USB is the interface not the media. You are talking about flash storage. They make IDE flash drives and USB magnetic disks. Fsck only checks for damage do to computer crashes when a disk is active. It does nothing to look for bad blocks. USB keys look cool, to whom? Not the chicks. Not normal people. Even most geeks think they are cheap.

Re:USB Key's

[Rolo Tomasi]

Addressing your points in the order presented ...

Sorry, can't quite follow you about the 10V thing. Yes, first generation flash chips had a Vpp of 12V, but that was about 15 years ago. The later chips all have integrated charge pumps to generate whatever voltages they need internally, so it's completely transparent to the circuit designer.

It's not about remapping the sectors, it's about distributing the wear so that all sectors get worn out equally. Once you get an error, the flash chip is probably already breathing its last. The point is that it takes so long for all sectors to wear out that it doesn't really matter.

The card I used had my calendar, contacts and database files on it, which got modified (i.e. written) a lot during a business day.

Flash isn't really suited for replacing HDDs in PCs, as I said, but this thread was about USB keys vs. hard disks for carrying around storage, and my point was just that flash cards or sticks or whatever are better suited for that than hard disks, and of course that the flash memories wearing out quickly is a bit of an urban legend.

Re:USB Key's

[Cyno01]

Put a keyring hole in it.

Re:USB Key's

[DonGar]

So, how well WOULD flash hold up as a replacement for a HD in a special purpose PC?

For example, if I setup a SolarPC with a 256 MB CompactFlash card as it's drive, and install Smoothwall, I end up with a linux firewall that has zero moving parts (no fan, no HD). However, it writes out log messages at LEAST once a minute, 24x7. And I want that logging, though the box can be configured to minimize it.

How long could I hope for such a box to last?

Remember that I'm crippling the write distribution algorithm by having a lot of the disk usage go to files that are static for long periods of time.

Re:USB Key's

[Anonymous Coward]

compact flash uses a disk and i have had mine fail for dropping it. I use several compact flash cards and while i love the tech it is not infallible.

Re:USB Key's

[Anonymous Coward]

Don't mind Tom, he's pathologically incapable of admitting fault. If he stops replying, consider it his appology for wasting your time.

Re:USB Key's

[Anonymous Coward]

Typical current flash parts specify a lifetime of million erase/program cycles. You still often see 100,000, but I assume anyone designing HDD replacement would go for the longer life. (Someone designing a cheap CF card for the consumer market might not, of course.)

You still see cycle lifetimes as low as 10,000 on some parts, especially flash integrated with microcontrollers that is intended to hold the program, and thus not be erased too terribly many times while in service.

Any decent flash file system will incorporate a wear levelling algorithm. So, the answer to your question is simply, "How long does it take to write the any sector a million times"?

Let's say you spit out 128 bytes of log message once a minute. 128 bytes/minute * 60 minutes * 24 hours * 365 days = 67 MB per year. So if 75% of your disk is locked up by those static files (and if your wear leveller isn't good enough to relocate the static sectors even when they don't change), you'll overwrite the rest of the card once per year. You're good for about a million years, then (by which time you'll need to upgrade anyway to play the just-released Doom 3). Feel free to turn up the logging. Adjust assumptions to match your own preferences.

Re:USB Key's

[Anonymous Coward]

yeah..... but.... that's a HDD. with rotating disks and all. so.... i mean, doesn't that kind of prove his point?

Re:USB Key's

[Jeff DeMaagd]

I am guessing you would want a system that works without a swap file and make sure you have enough RAM. For one, I think the power needed for sufficient RAM is less than the power needed to write to a drive, whether it be solid state or mechanical.

Re:USB Key's

[caouchouc]

Worst case, a 256mb CF card rated for 1 million writes being written to an average of twice a minute, with at least one cell being re-used on every single write, without a wear-leveller, will last about a year before the repeatedly re-used cell(s) die.

Just about every modern CF card has wear-levelling, though. They transparently remap their layout to spread writes out across all cells. The device they're attached to is presented with a virtual mapping of cells so that nothing appears to move; when in fact the boot block that looks to start at cell 0 may have actually moved to cell 42,000 the last time it was updated because those cells have experienced less wear.

You'll only go through one write cycle of the card's lifetime for every multiple of its capacity you write to the card. Even at 10% efficiency, this means a 256mb card will last until you've written approximately 25.6 petabytes to it in total across its lifetime.

I don't think I'd want to use CF for an active swap partition, but it should handle firewall logging duty just fine.

Re:USB Key's

[WNight]

Actually, yes, the remap sectors are free.

The remapping data moves, along with the actual data. The last thing you write is a translation table saying which virtual sectors the last so many sectors are pretending to be. Each remap sector points to the previous ones. You follow these back, reconstructing an idea of which sectors are which. Then you have enough information to return the sectors the OS really wants.

It takes a bit of CPU, and more reads than if the directory structure was always in the same place, but you can keep writes to a minimum, especially on any one area.

Then there's the idea of making the flash device with 5% more storage than advertised, using that to map into areas that die before the rest.

Re:USB Key's

[texaport]

Find me a USB key that holds 40GB for under $200

When 6000kbps DSL connections are as ubiquitous as the 300 million RJ11 jacks providing dialtone, you will be able to use a USB router that can "hold" every decent FTP site on the Internet. Until then, it'll cost you $150US per month to fill half your USB1.0 pipe with fast DSL.

--
Put that in your pipe and smoke...

Re:USB Key's

[haraldm]

Oh yes - ever run a multi-GB database off USB keys? Good idea.

Re:X-Box 2 anyone?

[ravenlock]

Ditch the complaining >:P the box looks neat to me, plus the added weight keeps it from falling to the floor when my kitten decides it's time to toy around with the controller wires :D ... although a miniature version in the spirit of PSOne would be nice.

2.5" drives

[FryGuy1013]

Hey, it's not the size that counts!

Re:2.5" drives

[AchmedHabib]

That's what small people say.

Re:2.5" drives

[jj00]

Hey, it's not the size that counts!

Actually, it is the size that counts. CNET News ran an article on the limits of hard drive size. They mentioned that the future business for these companies could be to work towards making hard drives that are smaller (and smaller, and smaller...)

Re:2.5" drives

[skookum]

The proper PC term these days is "libido-challenged".

Re:X-Box 2 anyone?

[Channard]

I keep mine on top of my digital box - I'd be too afraid to have it on a desk in case it did fall off and smash through to the floor below. And then through the Earth's crust, ending up somewhere in the Earth's core, being too hardy to even melt.

Power consumption

[Anonymous Coward]

...power consumption [is] apparently within acceptable tolerances

Unlike the eastern seaboard?

'sok. I'll get my own coat.

Re:Power consumption

[Anonymous Coward]

That's funny but except for New Jersey and Long Island, the eastern seaboard was fine. In Massachusetts and Connecticut the outage was confined to the western portions of the states.

Re:Power consumption

[evilviper]

I find it funny that my $50 UPS has better technology than the U.S. Power Grid.

Hell, all they need is heavy-duty brown-out protection to prevent these problems from spreading... I used to do simple little things like this is small electronics experiements by soldering a $0.05 resistor across the positive and negative leads, to prevent this kind of thing on a small scale.

The answer is obvious isn't it? Make power companies financially responsible for power outages, and within a year you will see the US power system suddenly updated to the best in the world... as if by magic.

Re:Power consumption

[Anonymous Coward]

This is so clueless I can't tell if it's flamebait or not. I'll bite. . .

I'm sure there will be some good technical articles on how powergrids work in the next few days. Do yourself a favor and read them.

Re:Power consumption

[Olathe]

...and within a year you will see the US power system suddenly cost the most in the world... as if consumers had to pay to fix it.

Re:Power consumption

[ZenShadow]

I do hope that you're referring to capacitors, not resistors...

Slashdot: large amounts of crappy information in really small doses. Usually.

--ZS

Re:Power consumption

[evilviper]

The problems with the power grid are stupid trivialities. It isn't going to cost you one dollar for them to make it reliable if they were motivated to do so.

Re:Power consumption

[Anonymous Coward]

"Slashdot: large amounts of crappy information in really small doses."

exactly... Why the hell would you connect a capacitor across an AC source? It would simply bypass your entire house, leading to a massive short before your circuit breaker tripped.

its even worse than a resistor.

Re:Power consumption

[evilviper]

I do hope that you're referring to capacitors, not resistors...

No, actually it was actually a complete mental fuck-up on my part... What I described actually prevents the load from being too _low_, not the opposite, which I wanted to discuss.

Wow!

[mrselfdestrukt]

What's impressive to me is that Tom's HG is still this fast , even with all of us checking out the drives. Tom must be the hardware king...

Very Nice

[Delifisek]

It will lead lots of different things
Like smaller desktop PC'S etc.

Re:Very Nice

[Pharmboy]

It will lead lots of different things
Like smaller desktop PC'S etc.

Yea, and hover cars. We have been promised smaller computers for 20 years now and the closest thing we have in mass production is a Dell Opti (not bad, I have one). 98% of all pci cards are still the same height as the original ISA (and S100 for that matter). The industry has put more energy into mod cases than smaller designs, and really we have to blame ourselves, since they make what we will buy. While having an aquarium built into your case has a certain degree of cool factor (inverse pun intended) we won't see smaller desktops until people DEMAND them.

We already have the technology and components to make very small and still powerful computers (ie: laptops) but people would rather spend the extra $300 (for the same power) on case mods or better speakers. Perhaps once LCD screens become standard equipment, smaller computers will become more in vogue. Until then, the size of a mid tower doesn't look so big compared to a 19" CRT. Personally I can't wait for computers the size of a CDROM drive, with midlevel+ power, I'm just not holding my breath.

Re:Very Nice

[Delifisek]

Hmmm I'm not talk about people who wants aquarium in his pc case.

I'm talking about small PC cases, or think how many of this 2.5 7200 rpm drives can fit 1u rack server ?

Also you point about size of PCI and AGP cards. So I'm asking is your VGA do same job your good old VLB cirrus logic card ? Or what. You may look at size of Realtek network cards. Or you may check Linux on chip systems. If Hitachi or Seagate can produce 73 gb SCSI 10.000 RPM 2.5 drive it will lead another options for the system builder.

And remember, if USA live another power shortage goverments may think raise of power prices.

Also I'm not asking to you hold your breath.

Re:Very Nice

[lawpoop]

We have been promised smaller computers for 20 years now and the closest thing we have in mass production is a Dell Opti...

I don't know, have you see these hand-held gadgets? Palm, Zaurus, et al?

Re:Very Nice

[EmagGeek]

The reason they can't make cases any smaller is because of heat dissipation. The smaller the case, the harder it is to move air through it. Slap a 60-watt processor in it and you have a huge heat problem.

What they need to do first is get rid of ribbon cables (fer chrissake!) and use serial 2-wire busses for EVERYTHING. Those cables are a) really messy and b) a hinderance to air flow.

Then, they need to stop making these horribly inefficient power supplies and start designing well-made supplies using good parts.

Then, they need to STOP using Linear Voltage Regulators for voltage stepdowns.

Of course, all of this costs money, so they won't do it. SATA is a step in the right direction, though.

If you want a really small computer, get an ITX case with one of those 800MHz or 1GHz VIA processors. They have them already that are the size of a CD-ROM drive - you just haven't looked hard enough :)

Re:Very Nice

[Pharmboy]

The reason they can't make cases any smaller is because of heat dissipation. The smaller the case, the harder it is to move air through it. Slap a 60-watt processor in it and you have a huge heat problem.

Ever see the older Apple Cubes?

If you want a really small computer, get an ITX case with one of those 800MHz or 1GHz VIA processors. They have them already that are the size of a CD-ROM drive - you just haven't looked hard enough :)

There was an article here on those, like $500 i think, about two months ago. I am just leary of VIA so far, and haven't seen much info on reliability. I have seen that performance is ~= a celeron at the same mhz.

Re:Very Nice

[CrazyClimber]

http://www.mini-itx.com/ ...and I'm pretty sure I have a link for a hover car around here somewhere.

Re:Very Nice

[default luser]

There was an article here on those, like $500 i think, about two months ago. I am just leary of VIA so far, and haven't seen much info on reliability. I have seen that performance is ~= a celeron at the same mhz.

You're thinking of the new Nemiah core VIA released a few months ago.

You see, VIA has been playing a BS war with this miniITX fad: their processors are inefficient, so they clock them extremely low and sell them as low-powered systems. Their 1GHz C3 produces the same amount of heat as a Pentium III Tualatin 800, with HALF the performance. The only reason VIA's offerings sell is because Intel won't offer you a LV Tualatin at a reasonable pricepoint, they stick them in Blade servers and jack up the price.

The C3 / Eden processor lacked such modern technology as OOOE and register renaming, or advanced branch predicition or a full-speed FPU. These were advertised as "features" that allowed the core to be simple and tiny, but what they didn't tell you was it was also very inefficient.

Recently, VIA announced the Nemiah core, which included sixth-generation features like a branch-prediction unit, OOOE and register renaming, plus a full-speed FPU. Sure, now it performs nearly as well as a Celeron, but WHAT A SURPRISE, POWER USAGE HAS CLIMED AS WELL. The fact is, it's still just a normal microprocessor architecture with no specific optimizations for power usage, and they can only sell it as "low power usage" because it delivers such low processing power

Honestly, if I were to point to the future of miniaturized computing, I would definitely say innovative designs like the Pentium M are the way to go. Compared to the Pentium M's highly-efficient design, VIA's Winchip-derrived "low-power" core is a dodgy hack.

Now all we need are system integrators to realize that stylish, small and cool-running desktops could be the key to restarting falling desktop sales, and we's see a whole slew of Pentium M offerings for the desktop. We don't have to be chained to space heaters, I wish Intel and integrators would figure this out.

Re:Very Nice

[Pharmboy]

Honestly, if I were to point to the future of miniaturized computing, I would definitely say innovative designs like the Pentium M are the way to go. Compared to the Pentium M's highly-efficient design, VIA's Winchip-derrived "low-power" core is a dodgy hack.

About 3 months ago, there was an article here about the new benchmarks of the M. I went and did some homework and found that in reality, the M was about as efficient (cycles vs. work done) as the PIII, and at similar speeds. Now it has the wireless and low power features, but it just proves how much the P4 is a POS. The M does look pretty good for miniture systems tho, I can't argue with that.

The IBM 970 (aka Apple G5) also has promise in this respect, but not on the Windows front. It is about the size of a celeron, uses less power than any similar powerful Intel chip (42W), and does significantly more work per cycle. They are talking about dual cpu laptops. I am just waiting for IBM to catch up with Apple's demand so it can start selling their own boxes with the 970 and GNU/Linux. They anticipate an empty 4x box to be 3500, pretty fine. Oh yea, it runs both 32 or 64 bit code faster than I64 or Opteron. And dual FPUs. And 1ghz bus. Some info here.

Re:Very Nice

[ManxStef]

We have been promised smaller computers for 20 years now and the closest thing we have in mass production is a Dell Opti (not bad, I have one).

How about the Gigabyte TA series?

I picked up a TA-1 for under 200UKP (just add 2.5" lappy drive and PC133 stick), and it's very nice - Eden 800MHz, 2xUSB (if you're after hi-speed USB2.0 you'll want the TA-2), 1xNIC, audio, PCMCIA slot, CD drive, etc. It even comes with a really nice mini BTC keyboard and optical mouse. Oh, and it's quiet too, unlike those little "vacuum cleaner"-like Shuttle boxes! Absolutely perfect for a little Linux box :)

Re:Very Nice

[WoTG]

There are plenty of smaller cases out there. Shuttle has done good business with their mini cases.

Besides, even regular PCs ARE smaller. I've got an old IBM 286 in the corner, it weighs at least twice as much as my current PC, and is probably twice the volume.

What they didn't touch on is...

[Colitis]

Durability. The faster the drive is spinning, the more strain on the bearings if your laptop (I'm assuming laptops are the biggest use for these right now) is moving around at all.

I'm afraid I'd rather a slow drive that'll take all sorts of abuse - using my laptop on the bus, shuffling it around on my lap, turning it around to show somebody something on the screen etc etc - than a fast one that isn't tough enough.

Re:What they didn't touch on is...

[cperciva]

On the other hand, smaller drives are likely to be more durable. Smaller platters means shorter arms, which means less "flapping" if you apply a shock to the drive.

As far as the "abuse" is concerned, I think head crashes are a greater danger than bearings dying.

Re:What they didn't touch on is...

[VCAGuy]

the more strain on the bearings if your laptop

Most (if not all) TravelStars use FDB technology now--basically they use oil in the place of little steel balls. It reduces power consumption and makes the drives much quieter--and more stable at speeds.

Re:What they didn't touch on is...

[msgmonkey]

Also since smaller platters have less mass dont they generate less heat?

Re:What they didn't touch on is...

[Anonymous Coward]

OT: Q: "Why do sound techs say 'check 1, 2'?" A: "Because on 3 you have to lift"

Re:What they didn't touch on is...

[Anonymous Coward]

Unfortunately although they will generate a little less heat, they will be more compact so the heat will be trapped closer to the heat sensitive parts. However they claim that this isn't a problem.

Re:Wow!

[ItsIllak]

That's just because most of the north eastern seaboard can't join in with the slashdotting.

40GB?

[phalse phace]

The article points out that with 40GB units suitable for server or desktop use,...

But 40GB isn't nearly enough for all that pr0n... erm... I mean all those illegal mp3's... erm... I mean, oh never mind.

Re:40GB?

[GNUALMAFUERTE]

my 40 gb wd drive isn't enough neither even for the list of things you didn't mention before : ) P.S: haha, good one.

Re:Wow!

[Pharmboy]

What's impressive to me is that Tom's HG is still this fast , even with all of us checking out the drives. Tom must be the hardware king...

What impresses me is that /. is still quoting Tom's HG here. The last several articles I have read have caused me to lose all respect for them. All fluff, no facts, lots of generalizations, and no real useful information. This has been discussed here previously, and many are like me, and no longer bothering reading Tom's reviews.

They used to be the king, but the king has no clothes. Sorry about the OT, but it's relevent enough if it keeps anyone else from wasting their time.

Meanwhile...

[RMH101]

...I just got a WD Raptor 10,000rpm SATA drive witha 5 year warranty (under 100UKP) for my desktop. Try and keep up, people!

Re:Meanwhile...

[Lord Ender]

Does it vibrate your entire house?

Re:Meanwhile...

[Anonymous Coward]

...I'm surrounded by 73GB 15,000 rpm U320 SCSI drives, 10,000rpm 146GB Fibre Channel drives, and controller boards of a 480GB drive... of course, I'm helping make these...

Re:Meanwhile...

[GoRK]

I assume you were making a joke, but I have a few of these too. They are amazingly quiet, at least compared to the 10K SCSI's I've used in the past. They are also really smoking for ATA drives ..

Re:Meanwhile...

[Anonymous Coward]

slowpoke, My 15K rpm ultra scsi 320 drive was included with my $750 dell server.

Re:Meanwhile...

[Anonymous Coward]

Yes, but they are SATA, not ATA.

Totally different beasts, and you can't really classify hard drives as only SCSI only ATA any more. That would give ATA far too much credit.

Re:Meanwhile...

[RMH101]

no, they're as quiet as something like a WD JB "Caviar" model. Really nice, and totally at home in a desktop. Very impressed so far: 5 year warranty and all...

slightly ot

[fyonn]

what I'm thinking might be interesting for doing servers on the cheap would be to do raid arrays with usb based drives. 2.5" drives are small and low powered enough to be powered completely via the usb bus, usb2 (well, the version of usb that does 480mbps) has enough bandwidth, if you dedicated one usb controller per drive and had your 2.5" drives each mounted in a small metal container with a ide2usb adaptor in it then you would have a nice, cheap raid array with easily removeable drives. usb controllers cost buttons and you could either do software raid or even a hardware controller which could be built for the purpose.

it could be alot cheaper than removeable scsi drives, the raiding software could mark the drives so that they can be put in in any order.

what do you folks think?

dave

Re:slightly ot

[patch-rustem]

what I'm thinking might be interesting for doing servers on the cheap would be to do raid arrays with usb based drives...

Put down the crack pipe, sir, and move away from the keyboard.

Re:slightly ot

[jtcampbell]

USB speed is highly dependant on the processing power of the "host" computer. This is one of the main things that differentiates it from firewire. So whilst you may be able to do this you may well need another processer per drive or something silly like that. This would obviously drive up the cost a lot.

Also the cost of all those IDE->USB converters and custom "USB raid controller" is likely to push the cost of the array above that of a (much better) SATA or even SCSI one.

You'd be much better using SATA. It's designed for this sort of thing.

Re:slightly ot

[bobintetley]

raid arrays with usb based drives...

Didn't someone do this already?

Re:slightly ot

[MarcQuadra]

Do you have any idea ho terrible USB is for pumping disk-storage type data? Watch your CPU-monitor while you write to a USB drive sometime and you'll never want to use one again.

You know how your system gets a bit 'laggy' when you write to a floppy? USB is worse. Hell, my mouse can't even keep up when I burn CDs over USB!

USB is for low-impact perhipherals like cameras, scanners, input devices, and pocket-drives. Firewire and SATA are much better suited for use with mass-storage devices. Already most motherboards are coming with FIreWire, and SATA is right around the corner.

Re:slightly ot

[fyonn]

my only issue with sata is that while it has tiny cables for data it still has a huge great big power cable. usb gives enough power over the cable for a 2.5" hd (is it enough for a 3.5" drive? not sure, it would be nice) so it's a single plug thats designed for easy addition and removal.

you could have little caddies with usb ports on and a 2x5.25" bay in a computer that could take 4 or 5 drives. I wouldn't really advocate a hardware raid adaptor as that would be expensive. software raid would do.

afaik ide2usb convertors really aren't very expensive. as for cpu power useage, well thats something I hadn't considered perhaps firewire instead then?

I'm glad you replied politely, it was only a brainstorming style idea. I think it would be nice to have harddrives that are easy to add and remove from a machine, even if it's only 1. external firewire/usb drives are alright but they kinda look messy on the top of a machine, it would be nice to have a handy bay in the computer where you can literally just slot a new drive in.

it's just a thought

dave

Re:slightly ot

[fyonn]

Firewire and SATA are much better suited for use with mass-storage devices.

alright then, firewire instead as thats another bus that transfers power along the same cable as data, which makes things very easy. it's the concept I'm putting out there, not necessarily the specific implementation.

dave

Re:slightly ot

[fyonn]

I know it's currently fashionable to flame first and speak later (if at all) but you don't have to be so rude. if you have something to say then join in the discussion but simply writing off the idea in that tone isn't helping anyone and makes you look like an arrogant prick.

usb was just an idea, firewire would still work too, and is probably better suited (if alittle more expensive) due to different cpu requirements. my point was to use a standard that allows power transfer along the same cable as data. I don't think that either usb2 or firewire gives enough power for 3.5" drives which is a shame (unless of course one makes sure it does, but I don't know if this would remain within standard)

as for usb naming, you're just trying to look high and mighty. I know what usb2 means to me, and I know what usb1.1 means, however with the recent spat about what usb2 is supposed to mean I thought I'd make it clear.

now scsi3 drives in removeable caddies "aren't cheap" and scsi cards are also "not cheap". usb/firewire controllers however are cheap, have alot of bandwidth and have predefined standards for mass storage devices on them (amongst other things). yes, 2.5" ide disks are more expensive that 3.5" die disks, but they are still alot cheaper than 3.5" scsi drives.

ide to usb or firewire adaptors are also pretty cheap. I'm not saying that my idea doesn't have it's flaws or that it's designed for the enterprise, it's clearly neither. however I do think it's an interesting idea that a few people might like and I don't think it's deserving of the ridicule that you're heaping on it but thats just what I think.

the other thing about it is that someone could implement this at home today with bits he could buy at a local computer shop. it might not be all that neat but it would work.

dave

Re:slightly ot

[jtcampbell]

If it's the size of the power connectors that's the problem then just use a connector with more conductors and route data and power over that. Like almost all removeable drive caddies you can buy now do.

Re:slightly ot

[fyonn]

then just use a connector with more conductors and route data and power over that

I was, it was called usb or firewire :)

I was thinking that it would be based on standards, that way it's cheaper and more likely to be accepted.

I've not got any plans on making this myself, I don't going to start carving bits of plastic out to make caddies etc. just putting an idea out there, if everyon hates it, oh well, I'll live. I still think it has some merit though.

dave

Re:slightly ot

[silas_moeckel]

USB 2.0 and enough bandwith for RAID? Lets see 480 Megabits a second divide by 8 for a whopping 60 Megabytes a second half duplex. I have single SCSI drives that do more than that. OK maybe it's reliability your after I dont think anybody has ever written multipath for USB though it could work with existing Mulipath drivers so we will give it the benifit of the doubt. So you need something with a pile of USB root hubs on a PCI-X card.

Now if you dont care about speed and just realibility maybe a little hardware mirror on a USB to IDE bridge chipset and pray the chipset doesent fail.

I'm sorry I'll keep my Fiberchannel and firewire USB is for cheap toys thats about as far as it can go without the ability to multiple initiators (someting every high speed external bus but USB has least that I can think of on a modern PC) flame on you USB zellots :) (I mean that in the nicest way)

Re:slightly ot

[Man Eating Duck]

That gave me a really good laugh! For some strange reason I find the idea of RAID'ing floppy drives hilarious. Beware of the monster RAID array of 125 USB floppy drives...

Re:slightly ot

[kingLatency]

USB is for low-impact perhipherals like cameras, scanners, input devices, and pocket-drives. It's not even well suited for some of these. Scanners often produce very high-resolution images that would be better transferred with FireWire and the same goes for digital cameras, especially as higher-resolution ones become more common. For most cameras, FireWire card readers are the only option since built-in FireWire is only on very expensive cameras.

Re:slightly ot

[MarcQuadra]

I'm sorry, I meant to say that it was for low-end CONSUMER-level perhipherals. Nobody who actually needs to get work done sticks with USB for long, it's flaky and slow.

Re:slightly ot

[evilviper]

you don't have to be so

Every one of us gets irritated and acts like an ass some times.

usb was just an idea, firewire would still work too

Firewire still isn't as fast as the normal interfaces, nor is it anywhere near as cheap.

my point was to use a standard that allows power transfer along the same cable as data.

That's nice. You could have pointed that out in the original post, and it would also be nice if I knew what reason/benefit you want power and data on a single cable. It makes sense for your portable drive, but a RAID array won't benefit from that in the slightest, as far as I can tell.

I don't think that either usb2 or firewire gives enough power for 3.5" drives

And in some cases, neither can even provide enough power for a single 2.5" HDD, so multiple drives sure isn't going to work very well.

however with the recent spat about what usb2 is supposed to mean I thought I'd make it clear.

My very point was, there is not, nor was there a "spat".

now scsi3 drives in removeable caddies "aren't cheap"

Fairly cheap actually... You don't actually need "caddies" per-se, since there are plenty of SCSI storage devices where you can just hot-swap bare drives... Getting those devices aren't inexpensive, but compared to a Firewire/USB2 to IDE adapter, you'd come out ahead with the SCSI option in price alone, not even considering the performance.

and scsi cards are also "not cheap"

For the number of drives you can put on them, SCSI cards are significantly cheaper than (even) IDE cards. Again, not even taking into account performance.

For the price you'd have to pay to get 3-4 firewire or USB2 cards, you could get a good SCSI adapter. Besides, each card typically only has one bus, so you would really need one card for each drive to maximize porformance, and even then, nowhere near the performance of SCSI (not trying to rub it in, just the facts).

yes, 2.5" ide disks are more expensive that 3.5" die disks, but they are still alot cheaper than 3.5" scsi drives.

As usual, not when you compare price/performance. As a matter of fact, you can get plenty of low-cost SCSI drives out there that aren't significantly more expensive than IDE for the capacity.

What annoys me about your whole scheme, is that you are making very weird price/performance choices for no particular reason I can see. I still don't understand why you'd want to have your system's HDDs running over firewire/USB2 at all...

the other thing about it is that someone could implement this at home today with bits he could buy at a local computer shop. it might not be all that neat but it would work.

That's what I'm wondering... Why in the heck would anybody want to? Even if you want 2.5" HDDs for power, heat, or noise reasons, then why not just hook them up to an IDE controller??? Better performance, less expensive, etc.

What exactly am I missing?

5600rpm?

[stevenrieder]

I believe that's 5400 rpm...

Re:5600rpm?

[x136]

You don't overclock your hard drives?

Re:5600rpm?

[Anonymous Coward]

Depends on how close you are tot he equator.

Re:5600rpm?

[evilviper]

Well, 5600 would technically be correct, as long as your Laptop continues to spin at 200RPM (and in the same direction as your hard drive, and while the Laptop is powered-on).

Re:5600rpm?

[Shaklee39]

No, you would have to be spinning in the opposite direction or else the spinning would be slower relative to yourself.

Re:5600rpm?

[PReDiToR]

/me cuts off the offending finger that made the typo.

/me proceeds to put it in a padded box and send it to the esteemable proofreader.

/me will not let it happen again.

Re:5600rpm?

[evilviper]

I didn't intend for you to be spinning with it.

The connoisseur's view

[vevva]

Much as I admire 40Gb in a 2.5" package I'm going to stick with my 100kb 8" floppies. I find the quiet modern drives don't have the same sound quality as the original 8" floppies. It's the whirring sound and the "kerchink" as I swap floppies 3 times per mp3 that adds depth to the listening experience. Nope.. these new fangled drives have no place in the system of a true connoisseur. (PS If any of your readers have replacement valves for a Collossus Mk 1 I would like to hear from you).

Re:The connoisseur's view

[Bushcat]

Floppy disk RAID system here for the cutting-edge Luddite.

Re:The connoisseur's view

[Pharmboy]

It's the whirring sound and the "kerchink" as I swap floppies 3 times per mp3 that adds depth to the listening experience.

If you love the "kerchink" of an older floppy drive, I have some older ZIP drives for you that you will LOVE, with a quite distict "CLICK" sound....price negociable or best offer. :D

Re:The connoisseur's view

[evilviper]

I certainly know the feeling... I never got over the love of hearing MFM hard drives.

Now I just sit right next to coffee makers, which have some audible similarites to the good old drives.

You know, I still have a couple sitting around... Maybe it's time to find my old 286, plug it in, watch as my electric bill goes through the roof, and spend about an hour recording the sounds of MFM drives. I may even make a CD out of it, and sell it to geeks trying to fill a void in their lives, caused by the disappearance of MFM drives.

Re:The connoisseur's view

[lucifuge31337]

You know, I still have a couple sitting around... Maybe it's time to find my old 286, plug it in, watch as my electric bill goes through the roof, and spend about an hour recording the sounds of MFM drives. I may even make a CD out of it, and sell it to geeks trying to fill a void in their lives, caused by the disappearance of MFM drives.

Let me know. I'll fire up the $2500 "gigabrick" (Seagate 1 gig, SCSI-1, full height that has enough spinning mass to make a full-tower rock back and forth on startup if you put in one of the top bays) and record that too.

Re:The connoisseur's view

[Anonymous Coward]

Hey, I remember that drive. I got one for $80 from a computer store - it was their old server hard disk. After putting Linux on it instead of Novell, it worked just fine for a few years. The thing that really got me was how hot that thing got, and ... well, it seemed very fast on seeks, probably because it had so many platters.

Re:The connoisseur's view

[SirDaShadow]

> as I swap floppies 3 times per mp3 that adds depth to the listening experience...

See...if you would've used Ogg vorbis, you may not need to swap floppies at all! (/ducks) :)

Re:The connoisseur's view

[wilper]

Yeah, they were nice.. I still have the disks that came with my IBM AT, they truly sang when in use. Later I upgraded to some seagate 40MB drives, that wouldn't spin up properly, but if you powercycled the machine after the first failure the disks would still be spinning on the second boot and things worked just fine.

Good disks btw, you could hear the disks when new e-mail arrived.

Re:The connoisseur's view

[Anonymous Coward]

what kind of connoisseur are you?!? Anyone that doesn't have tin-ears knows that digital lacks the "air", warmth, and "openess" of a true analog system.

any true connesseur would use punch cards...

Re:X-Box 2 anyone?

[Mikey-San]

Here's a series of questions for you:

1. Are you picking it up every five minutes for some reason?

2. Does it not fit into a backpack or duffel bag?

3. Is it too heavy to sit on your floor or entertainment shelf?

Most likely, the answers to these were "no", "no", and "no".

Why are we complaining about a console that, for 99% of the time, just sits there anyway? (And when you DO take it to a friend's house, it's not a bitch to carry in a backpack. I know--I do it.)

Bitch about the controllers all you like, but it's much smaller than a PC you might take to a LAN party.

Re:Wow!

[cperciva]

Tom must be the hardware king...

Either that, or Pair Networks know what they're doing.

Interesting, but what does it cost?

[dphoenix]

You have to remember .. the "cost" figure is strangely omitted anywhere from his review. People will pay for performance, but only within reason! However, inevitably, price will drop on these things and you will see smaller systems (tablets, tiny desktop pcs, consoles). It would be nice to make an even smaller media center PC using one of these.

Re:Interesting, but what does it cost?

[onyxruby]

Pricewatch shows venders with the 7200 RPM drive for $339 @ 60GB capacity. Although at the least the MSRP price *should* be included in the article.

Re:Interesting, but what does it cost?

[Anonymous Coward]

They were

next-->

going to put

next-->

in the cost

next-->

but ran out

next-->

of pages

Re:Interesting, but what does it cost?

[Anonymous Coward]

They already do... the toshiba satelite media centre laptop with built in tv tunner....

I'm pretty sure it uses the tosiba drive mentioned in the review....

Re:Interesting, but what does it cost?

[Wolfrider]

--Bah, unless I'm *really* interested in the testing process, I skim thru 2-3 pages and then skip right to the Conclusion at the end.

With 20k rpm scsi drives....

[Awptimus Prime]

With motors and bearings getting better, the biggest, baddest drives are pushing 25k rpms now. How does a 5k or 7.2k rpm drive keep up?

Not being the one to state the obvious, but since the rest of my computer is getting faster, why should I even consider a slower soltution because it's smaller? I mean, size would matter if the case of my computer had to be larger to house it (like a big drive in a little laptop), but since a hard drive is so small in comparison, I see no push for these to be adopted in workstations.

I do understand it being used in consoles and mp3 players. Just not in a workstation. Why must you say silly things in the headlines? :)

Re:With 20k rpm scsi drives....

[Anonymous Coward]

Excuse me? Please, tell us about these mythical 20K and 25K RPM drives. I dare you to post a link.

Re:With 20k rpm scsi drives....

[Alien Being]

only for very small values of K

Re:With 20k rpm scsi drives....

[Anonymous Coward]

Why I use 7200 RPM IDE HD instead of a 15000 RPM SCSI is because for me the price is a big issue.

Why 2.5" drives are getting important is because a lot of people are craving smaller and more quiet computers. See Mini-ITX, Micro-ATX, SFF, etc. I for one is really fed up with these behemoth tower cases with high noice levels.

Re:With 20k rpm scsi drives....

[MarcQuadra]

AFAIK the transfer speed of 10,000 and 15,000 SCSI drives isn't so far ahead of the 7,200 offerings, the improvement is in the latency. I have a SCSI system here at home and the combination of low-latency and tagged command-queing (buffering of future disk requests) lets applications start up lightning fast. Loading video and MP3 files is actually faster off the ATA drive though, as it has a higher raw transfer speed.

What would be really useful for workstations in this day and age are 16MB buffer 2.5" 5400 RPM disks running SATA with TCQ. An intelligent pre-buffering system to keep the 16MB buffer full of anticipated data would help the drives perform as well as 7200 RPM drives, and the lower heat generation and power requirements would boost corporate adoption. I'll bet that after a short while these drives could even be cheaper to produce, as there would be about 1/6th the raw materials needed; the only reason they cost more now is the lower volume.

Re:With 20k rpm scsi drives....

[amorsen]

An intelligent pre-buffering system to keep the 16MB buffer full of anticipated data would help the drives perform as well as 7200 RPM drives

Err no. At least not for reads. 16MB of buffer is not even half a second at full transfer speed. Besides, hard drive cache is at the wrong end of the cable. It is much better and cheaper to add those 16MB to the main memory. Perhaps if the cache was a few gigabytes, it might make a difference. It would still only cover a few percent of the drive, so the effect would most likely not be all that impressive.

16MB battery-backed write cache is a great help for synchronous writes though. Without the battery it is mostly useless if you care about your data: asynchronous writes are again much better kept in host memory than in a disk buffer. And synchronous writes have no use for cache that might get erased anytime, by definition.

Re:With 20k rpm scsi drives....

[toddestan]

Because 95% of the data on my computer is mp3's, videos, pictures, and that kind of stuff. Even the slowest of the slow 5400RPM drives can serve MP3's easily, and is able to serve compressed video too. I can see a fast boot drive for the OS, applications, swap files, scratch disk, etc. - but I hardly need 300GB of super fast storage.

Re:With 20k rpm scsi drives....

[MarcQuadra]

true, 16MB is only half a second of full-blast transfer, but the VAST majority of desktop disk-access is random small accesses, 32K here and 192K there sort of stuff. A better prefetching mechanism would help greatly for typical use, maybe have the drive read the entire track and put it into the buffer (and move the head to the next track) so the next access doesn't have to wait for data on said track to 'line up' again.

As for using the system's main memory, I'm all for it, someone needs to figure out how to have to OS profile how 'hot' sections of the filesystem are and prefetch all the indexes, directory info, and metadata into a smarter kind of disk cache.

Right now if I load Mozilla, 'dd' a 250MB file and re-load mozilla it doesn't hold moz in the cache. My OS should know that Moz is very 'hot' with several loads/day and it should keep all the metadata in the cache. I was also thinking that if we got really fancy, it could prefetch all the data in /usr/lib/mozilla and write a raw 'cache set' to swappable memory, loading a pre-organized directory of files from semi-swapped memory is probably faster than recursing the directories again and again, right?

As for battery-backed persistent caches, it could be made easier, I think, with a capacitor on the drive and some leftover inaccessible sectors at the end of the drive. When power cuts out the head drops back to the 'recover area' and starts dumping whatever's in the write buffer to the disk, when power is restored the first thing the drive does is check the 'recover area' for any 'dumped data', which it commits to disk (this time where it belongs) before your BIOS can even query the drive for booting. The capacitor would just buy you enough time to do that dump, and the drive could remain unplugged indefinitely without the risk of losing that data, unlike a battery-backed cache.

Re:With 20k rpm scsi drives....

[amorsen]

A better prefetching mechanism would help greatly for typical use,

That prefetching mechanism cannot be on the drive. The drive does not know whether a sector contains metadata, database locks, or video. Of course the drive can try to figure it out, but doing so would require vast amounts of CPU power and memory. Not something you want on a hard drive.

maybe have the drive read the entire track and put it into the buffer (and move the head to the next track) so the next access doesn't have to wait for data on said track to 'line up' again.

This is so last millenium. All disks do it already. The rest of the ideas have been proposed before. Some of them have even been implemented.

Re:With 20k rpm scsi drives....

[MarcQuadra]

alright, so it sounds like all the disk-end optimizations are tkaen care of. I am very aware that the drive has no idea how to intelligently prefetch data inside filesystems. My other thinking was much more aggressive CPU-end prefetch and caching, not this FIFO dumb-cache of most modern OSs. I want Linux to totally prefetch folder/file indexes that have been accessed most (up to 2/3 max disk cache size) in the background when it fires up. I wrote some scripts to recursively 'ls -l' my most commonly used folders and 'cat > /dev/null' my most commonly used files, and starting up apps is VERY much improved, even 'login' presents its 'password:' faster now, which was a major annoyance when I first booted the laptop.

Re:With 20k rpm scsi drives....

[Awptimus Prime]

Oops! I forgot they are peaking out at 15k and thought 20-25k had already come out. Perhaps I shouldn't have been such a hippy in my younger years.

My original post needs to be crushed into oblivion as to not contaminate search engines with it's mis-information. =)

Re:With 20k rpm scsi drives....

[Wolfrider]

> Right now if I load Mozilla, 'dd' a 250MB file and re-load mozilla it doesn't hold moz in the cache. My OS should know that Moz is very 'hot' with several loads/day and it should keep all the metadata in the cache. I was also thinking that if we got really fancy, it could prefetch all the data in /usr/lib/mozilla and write a raw 'cache set' to swappable memory, loading a pre-organized directory of files from semi-swapped memory is probably faster than recursing the directories again and again, right?

--What do you expect, man? Do you have 2GB of memory to play around with?

--Two things you could try:

1. I was going to recommend you set the "sticky bit" on various Mozilla files:
' chmod -cv a+t path/filename ' -- but found out that Linux ignores it. ( STICKY FILES - On older Unix systems, the sticky bit caused executable files to be hoarded in swap space. This feature is not useful on modern VM systems, and the Linux kernel ignores the sticky bit on files. Other kernels may use the sticky bit on files for system-defined purposes. On some systems, only the superuser can set the sticky bit on files. ) == http://www.die.net/doc/linux/man/man1/chmod.1.html

2. The other solution is to make a Ramdisk that's a little bit bigger than the size of the mozilla package uncompressed, and copy the mozilla package files to it. Then set your PATH variable to check the ramdisk 1st. You also might have to muck about w/ ldconfig, but that's beyond my scope.

--That's all I can think of ATM.

Bigger & faster, but not in my notebook

[Bushcat]

I routinely upgrade drives in my various notebooks, but I've discovered a drive in an external case can be much faster than swapping out the internal drive. To get maximum benefit out of the newer 7200rpm drives, one needs to use Mode 5, right? Do any current notebooks do that?

Hitachi have piles of info available on their drives here, and a discussion of 7200rpm drives here. The IBM legacy shines through.

id would rather

[m1chael]

want to see 2.5" cdrom drives (just pretend backwards compatibility wasnt an issue).

What about iPod hard drives?

[__aavhli5779]

These hard drives are truly remarkable in size. It makes me wonder what the deal is with the hard drive inside Apple's iPods. The largest one can fit 30 gb and a firewire controller into an enclosure the size of a deck of cards, and may indeed be one of the 2.5" hard drives reviewed in the article, or at least in the same class.

I can only imagine what an array of 40 of these bad-boys inside a rack enclosure could provide in terms of storage and redundancy. :)

Re:What about iPod hard drives?

[fr0dicus]

iPod hard drives are 1.8".... which is why the iPod is not at least 2.5" wide ;)

Re:What about iPod hard drives?

[Bushcat]

You're limited to 1.8" drives, which basically means Toshiba, topping out at 30GB with the MK3004GAH, a 4200rpm unit with slow access.

Re:What about iPod hard drives?

[Anonymous Coward]

also it costs a LOT more for a cute little apple product than for something that "just works"

Re:What about iPod hard drives?

[biggerboy]

That's ok. This is slashdot, where the pockets are bigger.

Re:What about iPod hard drives?

[Anonymous Coward]

Speed is relative, and the iPod hard drives are fast enough for what they're used for.

After all, when are users expected to saturate the top speed of the drive in an MP3 player? When copying mass amounts of music to the device, not when playing 320 KB/S MP3s. And 4200 RPM can keep up just fine with Firewire-400.

So it all works out nicely, and the physical size advantage that these drives have over 2.5" drives is of much greater value.

For Non-USians: The drive is 63.5 mm

[Anonymous Coward]

That's more likely to impress the ladies.

Re:X-Box 2 anyone?

[fr0dicus]

Try cost. Seen the price of 2.5" hard drives recently?

Laptops feel so slow...

[vudufixit]

And these drives with higher rotational speeds could alleviate the last severe bottleneck to good laptop performance.

Re:Laptops feel so slow...

[Pharmboy]

And these drives with higher rotational speeds could alleviate the last severe bottleneck to good laptop performance.

And they will not help power consumption a bit. Having power consumption "within tolorance" usually means on the high end of tolorance. For a desktop replacement, I agree wholeheartedly. For a notebook, I would prefer more effert put into acceptable speeds at half the power.

Re:Laptops feel so slow...

[vudufixit]

I'd like to see effort put into making batteries that last longer. Those methanol fuel cells have been two years away for five years now.

Re:X-Box 2 anyone?

[Anonymous Coward]

Why are we complaining about a console that, for 99% of the time, just sits there anyway?

Then I guess you wouldn't mind it to be twice its current size ? After all, if I don't need to move my console every 5 minutes, I guess I could use one the size of a refrigerator.

guess what, we don't all have a 500 m appartment to live in, and we don't all wear boxing gloves to play with our console.

How robust?

[mattr]

I always wondered if I turned on my laptop in a car should I maybe put the drive to sleep when going around tight curves or over bumpy roads. 'Course I no longer have that problem as the battery died, but does anybody know? Someone with knowledge of precession too.. would it be better to hold the drive vertical or horizontal? I thought smaller drives were stronger, does anyone know how much and whether the speedup for these drives hurts them in reliability?

I think I'd mostly like these to put a huge raid in a laptop or small case.

Re:How robust?

[bjschrock]

Actually, the smaller and faster the drives get, the more reliable they are (as long as component quality remains the same, etc). Also, depending on the use, the drives won't be spun up all the time. A drive with its heads unloaded can take a shock much greater than anything you could give it in your car without any trouble. And to answer your question: mount it horizontally, that will reduce the chance of HDI (Head-Disc Interaction, i.e. a crash) when you turn a sharp corner. A bumpy road would be worse on a drive than a tight curve if it is horizontal.

Re:How robust?

[Anonymous Coward]

I can answer this one from experience...I have an old Dell Inspiron 3000 with a 20GB IBM travelstar drive mounted vertically in the back of my truck as an .mp3 player (go Cajun!). Driving home from a friend's house in Vermont last thanksgiving, I went sliding off the icy road, off a 3 foot embankment and bounced across a farmer's frozen field for a few hundred feet before coming to a stop. Cypress Hill was still pumping from the speakers when I came to a stop. Everything in the cargo area had been "shaken & stirred". Surveying the damage, I found the laptop's CD-rom drive destroyed (cosmetically and mechanically), cracked connectors on the pcmcia serial port card, and a bent 1/8" to rca stereo adapter.

The drive continues to work. In 90+ degree F summer heat and -15 degree F winter cold. It's been running almost everyday for 3 years now.

Those drives are damn tough.

Oh please

[Gherald]

There are a lot of comments up there saying this will allow for smaller desktops, etc.

I don't think that is realistic. For the price you pay, 2.5" drives are horribly inefficient, and nowhere near as fast as 3.5" models.

Pretty much all 2.5" get used for now are notebooks and MP3 players.

Maybe as Mini PCs become more popular and mature these drives will get some use there. But this is hardly something to write home about.

Re:Oh please

[evilviper]

For the price you pay, 2.5" drives are horribly inefficient, and nowhere near as fast as 3.5" models.

With 2.5" hard drives at 7200RPMs, how could they possibly be called "nowhere near as fast as 3.5" models."

They whole point is that they ARE as fast... 3.5" HDDs have been stuck at 7200RPMs for some time now, and that is allowing for some very interesting competition.

Pretty much all 2.5" get used for now are notebooks and MP3 players.

Thank-You Mr. Obvious... The issue is not what they are used for now, but what they could be used for since they are now rivaling the speeds of larger hard drives, while still being far more energy effecient.

Maybe as Mini PCs become more popular and mature these drives will get some use there. But this is hardly something to write home about.

Maybe as these new cancer drugs become more popular, more people's lives will be saved. But this is hardly something to write home about.

Thank you for comming to /. and spreading your incredible lack of foresight, and sheepish consumer-istic attitude.

Re:Oh please

[Alien Being]

With 2.5" hard drives at 7200RPMs, how could they possibly be called "nowhere near as fast as 3.5" models."

Track seek times should/could be better on a smaller disk. But at the same rpm, number of heads and bit density data transfer would be slower due to the lower linear velocity.

They whole point is that they ARE as fast... 3.5" HDDs have been stuck at 7200RPMs for some time now

I've seen speeds as high as 15k.

Re:Oh please

[MarcQuadra]

the only reason they perform so badly is because there's no decent buffering in ATA drives. With Tagged Command Queing and big prefetch and write buffers they could perform well at all but transferring very large data files (not a common case on most desktops, BTW).

Also, why don't modern operating systems aggressively prefetch file system metadata and keep it in cache? It seems to me that with most systems having more than adequate RAM it would make sense to keep the entire directory listing and metadata cached. I think Reiser4 has a method to determine how 'hot' a directory is based on how often it changes, there should be a way to tell the kernel you want to aggressively prefetch indexes for all objects on '/dev/hd?' with 'heat=>2'.

Re:Oh please

[evilviper]

I've seen speeds as high as 15k.

Not in IDE drives... These 2.5" drives are all IDE. When they compare 2.5" SCSI drives, then we'll see how that works out.

Re:Oh please

[bjschrock]

For the price you pay, 2.5" drives are horribly inefficient, and nowhere near as fast as 3.5" models.

Talking about the models reviewed, yes, but that's soon to change, very soon. Seagate's Small Form Factor drives will be around next year. At 10,000 rpm and with a U320 SCSI, Fibre Channel, or Serial-Attached SCSI interface, they're as fast or faster than most of the 3.5" drives out there. The platters in the 3.5" enterprise drives are as small as the ones in the 2.5" anyway, and you'll (almost) be able to fit 4 2.5" drives in the space of one 3.5".

Re:Oh please

[Gherald]

These aren't servers with redundant power supplies we're talking about. The more you prefetch or write cache, the more you risk loosing data.

Re:Oh please

[Gherald]

Thank you for comming to /. and spreading your incredible lack of foresight, and sheepish consumer-istic attitude.

Lack of foresight? Its more like a lack of enthusiasm.

2.5" drives are improving at almost the same rate 3.5" drives are.

BFD

Re:Oh please

[MarcQuadra]

prefetching is by nature read-only, so we can't get hurt there, writing while losing power is an easy fix that nobody's implemented. Really.

Re:Oh please

[Gherald]

When there is a way to commonly implement that, tell me. THAT would be news. These 2.5" drives are not.

Re:Oh please

[Alien Being]

What a jerk. If you want to differentiate IDE from SCSI then go back and correct your own post, not mine.

Re:Oh please

[Anonymous Coward]

And as speeds get higher, platter size must be reduced to improve stability (less wobble) and lower power consumption. Most of those 15000 RPM SCSI drives have platter sizes comparable to the 2.5" laptop drives, and much smaller than your average desktop. (also, your 7200 desktop drive already has a noticably smaller platter than a 5400 drive, for the same reason)

The only reason desptop ATA drives are stuck at 3.5 " platters is it allows manufactures to have higher capacity to price ratio. There is no performance benefit to large platters, and current drives will certainly have to decrease in size before 10 000 rpm is possible.

Re:Oh please

[evilviper]

What a jerk.

It's all in your head... I wasn't insulting you or anything like that, just pointing out the difference.

If you want to differentiate IDE from SCSI then go back and correct your own post, not mine.

I can't list every single factor in everything I happen to say. I thought it would be a natural assumption that I wasn't talking about SCSI hard drives.

I also didn't differentiate between hard drives of different heights, voltages, operating environment, et al. Even though all those factors could make a great difference, it's usually safe to assume that someone is talking about the most common examples, unless otherwise noted... As much as I'd like it to be otherwise, SCSI isn't very common at all, so I didn't think it required pointing out that single exception.

Re:X-Box 2 anyone?

[kinnell]

plus the added weight keeps it from falling to the floor

It also stops it from blowing away in a strong breeze

Re:Wow!

[AchmedHabib]

Yes, because you can always use another harddrive, when you have upgraded the motherboard, cpu RAM, graphics, etc. all there is left is to buy a harddrive when you get that urge to buy some new hardware. At least that's what happens to me. At least that's what I do, buy a new HD and sell the "old" one to my friends. They have a never ending cheap source of slightly used harddrives.

"within acceptable tolerances."

[PurplePhase]

You mean 1 Db, 1 Calorie/square inch surface dissipation, and 1 Watt usage, all under maximum load in continuous use, along with 5ns seek and 1GB/s non-burst read/write? I find those to be acceptable levels for hard drives.

No, Seriously (except for missing all the other contingencies: physical size, cost, interface, GB capacity etc.). Until that happens, I'm just putting up with a (crappy?) industry design.

Similar to automobiles.

8-PP

smaller != better

[meshko]

Even full size harddrives have gotten much less reliable lately. I assume this is because the data density keeps growing. I would rather buy a hard drive which is slightly bigger (I guess it would have to be more platters because making the radius bigger will make seek times longer?) but will last for more than 2 years.

Re:smaller != better

[Anonymous Coward]

All i know is my laptop has a 4 year warantee on the hard drive, much more than you can get on any consumer drive (other than the 10000rpm raptor)

Smaller sizes actually increase reliablility, most enterprise-class drives have platters comparable in size to 2.5" laptop drives

Re:USB Keys with HDs - already happening

[ioliver]

Cornice Inc's 1" sized and 1.5G capabity Storage Element (that's a HD to you and me!) is going in to consumer devices such as cameras, and now also in to a USB "key".

http://www.hpcwire.com/dsstar/03/0610/106016.htm l

A couple of points...

[Anonymous Coward]

I am a storage consultant so I kinda know what I'm talking about here (just thought I'd get that in before I get slagged off) and assuming that you're not totally joking...

1) The technology used within USB type memory keys is only good for about 10000 write operations max.

2) They are very expensive

3) I don't see any USB -> Fibre Channel converters and none of my suppliers have them on their hardware roadmaps (can't think why)

4) They are staggeringly slow, even if you RAIDed a thousand of the buggers.

5) If anyone took one of these keys into a datacentre in which I was responsible for the storage, I would do some painfully biological things to them.

6) In modern datacentres the mass storage (and quite offen the local system disks as well) are supplied from a consolidated disk array which is hung off a fibrechannel network almost nobody who is anybody does JBOD for mass storage any more.

7) RAID shouldn't ever be controlled by software for serious users

8) can't be arsed to go on, but you get the general idea...

Re:A couple of points...

[fyonn]

1) I mean usb interfaced ide drives, not usb keys.
2) see 1
3) fibre channel is getting way too expensive, this is all about being on the cheap
4) see 1
5) see 1
6) yeah, I've worked in a datacentre, I know. this is not about being for enterprise storage and such but just a cheapie method for home use. people buy usb and firewire drives after all, this is expanding on that a little
7) for business use I agree with you, for home use I would say that it is acceptable depending on the situation
8) :)

people seem to think I'm on crack with this idea (re the first reply to my comment) without really listening to what my idea is. okay, it's not one of my best ideas but I do think that it has some merit, even if someone only created a 3.5" bay that took a single usb/firewire HD in a caddy. it could be pretty cheap and makes for easy hot-swappable storage. also if whoever you take it to doesn't have the matching caddy then they can just hook up with a normal usb cable/firewire cable depending on what tech you decide on.

come on, ppl, I've got +5 interesting on this, someone give me a positive reply please :)

dave

dave

Re:A couple of points...

[Anonymous Coward]

RAID shouldn't ever be controlled by software for serious users

Don't know what you define as serious. The hardware RAID controllers I used work better as software RAID. Reliability, easy of setup, and performance are all better. I just switched that last RAID from hardware to software. Now, understand, this is IDE RAID. We have a budget. I also needed to fit 8 disks and 2 CPUs in 4U. These computer travel the world. Software RAID isn't as bad as it used to be. It doesn't make a lot of sence having very complex programs running in firmware. As far as performance, CPU time is cheap these days. Anyway, one solution isn't perfect for everyone. A serious admin knows that. You need to consider all the alternatives and not ingore any.

Re:A couple of points...

[dissy]

Ok, here's a positive reply.
Ive started on this very thing myself, using firewire however.

Unfortunatly I started getting all sorts of ideas and tacked them on, and now 'cheapie' is out of the question.

I built a system with 4 firewire buses just for disks.
I also chose not to power the disks from the firewire bus (explained bellow why)

The master plan is to have 6 firewire buses (two groups of three, and they only had dual bus cards, thus why i have 4 ports now.. One is not used yet, and will be used with the 3rd dual fw card i do not yet have)
These three buses connect to firewire hubs.
Then, you connect three disks per hub.

Now is the confusing part.
On bus A, you have 3 disks. These are disk A from 3 different raid-5 groups.
Bus B has 3 disks which are disk B of the same raid-5 groups. and so on for C.

Then you setup a raid 5 group out of just those 3 disks. Which in this case gives me 3 groups, and with the 3rd firewire card (bus 5 and 6) this concept will double.

Then you use LVM to link all of the raid-5 groups together into one big volume.

Reasonings:

If any one firewire bus failed (or was unplugged) only ONE disk from each raid-5 group is offline. Raid-5 can survive this.

I dont use power over firewire because a) the PCs supply can not handle all of that, and b) I now have 3 power supplys, each one chained to the disks in the same order as the firewire buses. This way if a power supply failed, it only takes down one disk per raid-5 group, and again raid can survive that.

Firewire is multi-host (IE you can have more than one host controller on the same bus) so with two computers on the bus, doing heartbeat monitoring over a serial link between them, if a computer failed, the other can pick up the disks on the bus and continue file serving.

Using LVM to link the raid 5 groups together means after i start getting disk failures 4-5 years down the road (well, hopefully that long) and it starts to get hard/expensive to find disks of the size i am currently using, I can move the data off the raid-5 group to unused space, and decomition that group. Then it can be pulled off the bus, and replaced with current newer disks which ideally will be much higher capasity, without replacing ALL the disks in my array (as would be the case with a single raid-5 array of all the disks)
Then you recycle the failing disk, and have two disks spare to use for other machines (IE a mirror to boot a new machine off of, spare single disks, etc)

Some links you may find interesting:

- http://www.fwdepot.com/
Best source of firewire controllers, bridgeboards (firewire -> IDE, firewire -> scsi, usb->ide, enclosures, clamshells you mentioned, etc etc)

- http://evms.sourceforge.net/
EVMS = Enterprise Volume Management System. Linux software that lets you manage raid, lvm, clustering, etc all from a server setup. Comes with cli, curses, and X11 interfaces. Not quite 100% there yet, and still has a couple problems for enterprise use, but almost all of them are related to the 2.4 kernel and promised to be fixed when 2.6 is out/stable (and in the past month very well could have been, i havent been keeping up)

(Please please dont slashdot my poor little file server here!
If anyone would like to mirror, its ok with me. This is a p2 200 and will die if more than a few peope hit it at once)

- http://photo.brokensphere.net/index.cgi?mode=view& album=home/fileserver
(Watch for slashdot injected spaces!)
This is pictures of some of the parts at the start of my project.
Havent added new picts yet, nor had much time to work on it.
These picts show the disks all on one bus and interlinked, which is not good for speed, but I was just testing the EVMS software at the time.

In the end, I plan to make my own case, which may be a sheet of half-inch think wood screwed into a wall, with disks hanging on it as so their tops face out, and a plexigl

Re:A couple of points...

[fyonn]

thank you :) yay, a positive reply. well done, your project looks quite interesting. you've gone to alot more effort than I have. I've just thought about this a bit and lamented it's none existence, you've actually done something.

like many people, I used to have a harddisk in one of those removeable caddies for taking stuff to friends, usually pirated software, back when I was into that sort of thing. the problem with that was that it was big, clunky, not hot swappable and different caddy manufacturers all had their own connectors, well, quite a few used the big 'ole centronics ones, but they all put it at a slightly different offset etc :)

when I discovered usb and firewire hd's I thought it was a great innovation. everyone has the same connectors on their mobo and you could just take a drive around, plug it in and use it. problem being of course that you had to power the drives seperately which was clunky, and they were always external which made things untidy. I always though it was a shame that no-one ever made an internal caddy for usb or firewire drives. as I said in previous posts, a 2.5" hd can be bus powered by either bus format and it would be very neat. the thing in the machine could line the caddy up, as you push it in, with the plug at the back and as you push it in, the plug makes contact with the socket and bob's your uncle. and it's hot swappable by design.

if the caddies look good then it could be a zip disk replacement, esp as the other person doesn;t need a matching caddy, they can just use the normal cable.

after that I thought that it could be a great way to do cheap ide disk raid or even JBOD. if the disks were marked then it wouldn't matter what order they were put in and you could do with them what you like. after all, some machines can even boot off firewire.

dave

Re:A couple of points...

[Chaset]

Along those lines, VST (now SmartDisk) had a 'tower' of sorts where you can plug in up to 6 (or so) of their 2.5" firewire chassis. The bottom slot was for a Powerbook style battery so that the whole thing can run off of battery. It was designed to be RAIDed. I thought it was a cool idea, but unfortunately, they don't seem to sell it any more.

Re:A couple of points...

[Anonymous Coward]

I would do some painfully biological things to them.

I don't think you said what you meant to say... 'Painfully biological'? I mean, I know biology can be painful, I had to take a year of it, but come on...

Re:Wow!

[PReDiToR]

That was my main worry in submitting the article, I still check out Tom's every day, but it seems that it heads more towards the gaming market every month.

I think this proves that /. doesn't judge the source, onlly the article submitted.

Right on!

[uradu]

They've sold out all the way, the site has gone all commercial. I stopped frequenting them regularly when they started reviewing movies and games. It seems most of their "serious" hardware and new tech articles hang around forever without any updates.

Re:Wow!

[generic-man]

All the real hardware geeks are in Michigan, which won't be fully on-line until late Sunday.

Re:Wow!

[Pharmboy]

I think this proves that /. doesn't judge the source, onlly the article submitted.

In all fairness, not every Toms article sucks, but the only ones I see now are the ones that are posted here on /. I don't go there on my own anymore. The ones that have been posted as articles here on /. lately have not been up to snuff, and instead tend to be alot of words that cause you to go from page to page and look at ads. Toms is not alone in this fact, since the shakeout of 2000+, lots of sites have gone downhill or out of business since the investors and business owners are finally insisting on making money with the sites. IMHO, we seem to still have more tech/hardware sites than the market will support, thus the quality is not great. ./ itself is one of the few exceptions. Many bitch about the lamers and flamers here, but I have been on /. for quite some time, and the % of noise hasn't changed that much, mainly due to moderation. Now, I'm not saying /. is perfect, just has stayed stable in the % of lamers. ;D

Is that a hard di_k in your pocket?

[DickBreath]

Don't you really mean...

Is that a hard di_k in your pocket?



Mmmmmm. My favorite kind.

HP has plans for 2.5" server drives

[Anonymous Coward]

According to a server roadmap that HP presented at HP World this week, HP is planning to standardize on 2.5" drives in new Proliant servers in the 2004-2005 timeframe. The reason that was given is that with platter sizes getting so large on 3.5" drives and leading to larger drive capacities, customers want smaller drives for their servers for performance purposes. By switching to 2.5" drives, HP can offer more drive spindles in the same space that current 3.5" drives reside in. I didn't think to ask the presenter about drive speeds, however, since it was an end-of-day presentation, but I'm sure the gains from increased spindle counts don't come anywhere close to making up for the slower RPM's of the current and near-term 2.5" drives. BTW, this was an NDA presentation, thus the reason for the AC posting.

Re:HP has plans for 2.5" server drives

[amorsen]

Today's SCSI drives are almost exclusively 2.5" and smaller. They just come in big packages... That is, internally the platters and electronics are easily small enough to fit in a 2.5" drive (taller than notebook drives admittedly).

It will be interesting to see if 2.5" server drives will arrive as parallel SCSI, serial SCSI, or SATA first... Either way, 2.5" 15k drives are just around the corner. They will not work in a normal notebook though; it would melt.

Re:HP has plans for 2.5" server drives

[Anonymous Coward]

Looking back over my notes from the presentation, it looks like serial SCSI will be the standard SCSI connection on Proliant 500-700 series servers in the 2004-2005 timeframe. It looks like the 2.5" drives are coming quicker than that, early to mid 2004 according to my notes. They have no plans for SATA on this line of servers, so I'd expect the 2.5" server drives would first arrive as parallel SCSI but be designed so that they can easily switch to serial SCSI in manufacturing once they standardize on that technology.

New filesystem for USB Keys?

[DickBreath]

Perhaps what we need, both for USB keys, and for CD-RW, DVD-RW, is a new filesystem.

The goal of such a filesystem would be to minimize the re-use of any given sector. It would be a no-no to have a single superblock that never moves, for example. Even if the disk is mostly empty, but I keep re-writing the same text file, the filesystem would try to uniformly manage the use of the sectors on the disk in such a way that each subsequent rewrite of the small file goes into a different set of sectors.

Re:New filesystem for USB Keys?

[Russ Steffen]

Why a new file system just for that? UDF already handles wear leveling for CD-RW and DVD-RW. And there is JFFS for flash memory. And, those are even necessary for Compact Flash, SD, Smart Media, or those USB flash drives as they incorporate hardware wear leveling. Those will last the same length no matter what file system you use.

Re:USB Keys with HDs - already happening

[b-baggins]

Transferring 1.5 GB over USB?!? I think not. Firewire 800 maybe...

Breakfast?

[Anonymous Coward]

Wow. What did you have for breakfast this morning? Quaker Instant Bitch?

Re:Breakfast?

[evilviper]

Incredibly moronic comments that make no sense at all, yet get moderated up to +5 really get under my skin.

If he was mistaken, or had some facts wrong, not a big deal... But his entire idea made absolutely no sense what-so-ever, no matter how you look at it.

Autonomous Computing?

[fygment]

Right our university has a mix of Unix and Win2K computers with different settings as you go from dept to dept. All I/O is to floppy or some rather rare CD RW.

It would be nice if all the university computers were without a HD. A student would be issued a 2.5". To log on, insert the drive into a bay (like a 2.5" slot or something) on a computer. Voila the computer boots to your personal settings with all I/O going to your drive. Done, pull the drive and walk away. Any computer you use will always give you the same environment.

Just a thought that seems closer with these size drives.

Re:Autonomous Computing?

[derrith]

the problem with using 2.5" drives for user customised machines is that EVERY machine has to be exactly the same as next one. Well, not exactly, cpu and ram can be different. However, Motherboard, video card, whatever peripherals, sound cards (if any) would need to be exactly the same as win2k and unix do not take so kindly to hardware chaning on them from one bootup to another. what may work is a FAT32 home partition on the unix boxen which will just pop up as another drive on the 2k boxen. just my thoughts on the matter.

Re:Autonomous Computing?

[toddestan]

What if you used an OS like Knoppix on the drive? Knoppix boots into just about everything but the most bizarre hardware out there.

Re:Autonomous Computing?

[fygment]

Ah. I guess I was just so inspired by Knoppix but didn't clutch in on the importance of the hardware detection. Thanks.

A Cheetah is Faster and it has 2.5 inch platters

[cheetah]

I have one of those also, but I have had a 15Krpm Seagate cheetah for over a year now. The cheetah smokes the WD drive, but that is to be expected. The WD is about on par with a 10Krpm scsi drive from last year( performance and price ). Still with Sata they have removed most of the difference between Scsi and Ide( to be fair Sata is more like scsi than it is like Ide, but I won't complain :). Oh and to keep this at least a little bit on topic the platters in my cheetah are only 2.5 inches. They made them smaller so they could improve seek times and reduce the power draw when running at 15Krpm.

Below cost

[SuperKendall]

You might have a point except that if you try to buy a Toshiba HD all by itself, it ends up costing about the same as the iPod that contains the same model...

Re:Wow!

[sharkey]

Either that, or Pair Networks know what they're doing.

They must, to stand up to Slashdot linking to a site that measures content in pages per sentence, rather that sentences per page.

Price

[phorm]

I do understand it being used in consoles and mp3 players. Just not in a workstation.

Depending if they go down enough in price (and after a few years, they likely shall). If they become common - we could see something to the extent of a hot-swappable minidrive enclosure for microdrive. Basically it would be like floppies except a lot bigger and of course using hard-disk architecture.

Really convenient when, say, you're a tech or something else that might involve transferring large amounts of files (of course, DVD's are 4.3GB and drives getting cheaper, but not really so convenient to rewrite).

Re:Price

[Awptimus Prime]

Oh. I never thought of it that way. You could have a ton of those things shoved into the front of even a small PC case.

How about a portable raid array that plugs into the front of your machine w/ multiple drives braced together as a cartridge? Get the technology up to speed a little and running 4 in a raid0 array could be a pleasing experience.

Cost and size seems to have prevented a lot of neat things from happening so far.

Thanks for the enlightenment.

Bigger != better

[phorm]

Smaller drives would result in less worry about "wobble" (of the drive platters) in the case of shock, and less travel-distance for the needle. Now the real question is, are drives dying to demagnetism (caused by greater cluster density) or just shitty design? The lost one I had frag out on me (30GB maxtor) overheated to death, and others have been issues with the needle, so I'm guessing that the less movement that is required of mechanical components the better!

become smaller?, o yhea, look at my Full Tower ...

[GNUALMAFUERTE]

Become smaller?, why should they?

Laptops are cute (i have an _old_ travelmate : ), doesn't have enough power yet ... (may be if you can afford the cost of a Crouse ... )

But at home i have my 91 cm Full-Tower 5 1/4, 2 3/12 Case!

Small Computers Sucks
Let's go back to 5 1/4 '' 2 GB Big Foot Hard Drives :)

Bah...

[El Camino SS]

Nothing compares to the gentle, lilting, swish-swixh noise as your Commodore spends fifteen minutes booting a text type adventure from a audio cassette. It's as calming as a whale song.

Don't forget to be kind and rewind.

NOW THAT'S STORAGE!

Re:Bah...

[Anonymous Coward]

You may have missed the joke... do a search for "click of death"

Here's my iPod question...

[El Camino SS]

Does the sleek design and trim size prevent your kitchen table from wobbling after you accidentally drop it for the first time, or the second time?

(Personally, carrying around a $300+ hard drive scares the living crap out of old butterfingers here.)

Re:USB Keys with HDs - already happening

[ioliver]

I use a full HD snapshot program http://www.drivesnapshot.com/en/index.htm to make full images of my notebook HD onto an external USB2 HD. Agonisingly slow with USB1.1 but about a minute per gig (before compression) with USB2. Ian

A current use...

[Anonymous Coward]

We are already currently using these in our blade servers. I love these little things, they make maintaining a blade as easy as a full size enterprise server.

Re:slightly ot - use firewire

[DonGar]

Switch from USB to firewire (which ups the cost), and it sounds great to me! The concept is good. In fact, if you ever build something like this, please post an article to SlashDot about how it turns out. I might even try something similar myself, next time I do a major server upgrade.

That said, USB is just a clunky protocol, and would involve a huge CPU drain to run heavily, especially with multiple buses. Also, because USB is known as being slow, all the people along the way were less worried about being efficient. This means the drivers, the USB to IDE hardware, the USB adapter, everything.

I'd really be suprised if anyone has every examined and optimized drivers (for any OS) to handle multiple adapters well.

Firewire is a smarter protocol, there is less CPU involvement, and it known as a 'higher-end' solution, so most of the people involved would have tried harder to perform well.

I would try it with just two drives to see if the raid software would really work with external devices, especially with live connects and disconnects. I'd be suprised if any raid software could handle having devices created and destroyed under them (as compared to just going into an error status) without some issues, but there is a good chance that the fixes would just be minor tweaks.

Re: 2.5" platters in 3.5" drives

[booch]

Yes, I just read that in this excellent hard drive reference. There's a chart showing that 10K drives are 3.0" and 15K drives are 2.5". It also explains why smaller platters are better.

floppy raid

[chocolatetrumpet]

It's been done.. usb floppy raid!

Re:Very Nice-Core memory.

[Anonymous Coward]

One we can get this done in tight bundles. It could be a replacement for hard drives. Theoretically we could criss-cross a present day platter with a R/W array. And don't forget quantumn dots in the future

Laptop upgrade

[LittleLebowskiUrbanA]

I was thinking of upgrading my laptop's hard drive to a 7200rpm 8MB cache 60 gig but I couldn't tell if Tom's was testing the one w/ the 8MB cache.

Whoops- correction

[caouchouc]

Ugh... I skipped right over "giga".

That's terabytes, not petabytes. Sorry.

Big for a reason

[Anonymous Coward]

Computers are big because big things are easier to handle. Imagine having a wrist-watch sized PC. Every DIY guy will now how to buy a robot attached to a microscope to replace that microchip sized PCI card.

iPod does this today

[Anonymous Coward]

Get an iPod or any of the other portable mp3 players out there today and you can use it just like you described, especially with the new dock.

Re:A Cheetah is Faster and it has 2.5 inch platter

[RMH101]

Since I bought these pills I saw advertised in my Hotmail, I now have a 13" dick and can beat people to death with it....!

Re:slightly ot - use firewire

[fyonn]

yeah, the general consensus is that usb isn;t the best idea, and as it happens I have a single 2.5" bus powered firewire disk, not usb. (which freebsd doesn't work correctly with, which is annoying :)

I hadn;t realised that usb had such a bad rep but firewire is just as good, although the controllers might be a tad more expensive. only thing is that atm I really can't afford to play around like that atm, I've just had a lappie pinched so

one day perhaps, mind you, the other guy seems to be doing the idea well enough.

dave

Re:X-Box 2 anyone?

[ravenlock]

And it's great for weight-lifting too!