Seagate Launches Hybrid SSD Hard Drive

MojoKid writes "Though there has been some noise in recent years about hybrid storage, it really hasn't made a significant impact on in the market. Seagate is taking another stab at the technology and launched the Momentus XT 2.5-inch hard drive that mates 4GB of flash storage with traditional spinning media in an attempt to bridge the gap between hard drives and SSDs. Seagate claims the Momentus XT can offer the same kind of enhanced user experience as an SSD, but with the capacity and cost of a traditional hard drive. That's a pretty tall order, but the numbers look promising, at least compared to current traditional notebook hard drives."

Comments

4GB?

[bbqsrc]

4GB is very restrictive. I had the first EeePC, the 701, with the 4GB SSD. Talk about micromanage the OS... You can't even install Windows 7 onto the 4GB SSD and put the rest of your data on the HDD, as it's too small! What's the point?

Re:4GB?

[WrongSizeGlass]

The cache is 4GB, the drive is up to a 500 GB 'traditional' drive.

Re:4GB?

[bbqsrc]

I see. Then I must leave.

Re:4GB?

[Anonymous Coward]

Well, why do you think they call it a "waist"????

Re:4GB?

[Yvan256]

Please leave your nerd card on the desk on your way out.

Re:4GB?

[Rallias Ubernerd]

If you must leave, can i have your nerd club official library card?

Re:4GB?

[Hatta]

That's bigger than any other drive for your XT. That is why it's called the Momentus XT, right?

Re:4GB?

[Walterk]

The cache is 4GB, the drive is up to a 500 GB 'traditional' drive.

So you just use an SSD as cache, writing to it very often, exactly what they're really bad at?

Re:4GB?

[InfiniteWisdom]

Yes

Re:4GB?

[TrisexualPuppy]

Solid state drives are flash-based and may never have the price:GB ratio that mechanical drives have. Methinks that even when 1TB SSDs become relatively cheap, mechanical HDDs will still reign in price and long-term reliability for heavy data usage. The deal with this hybrid product is that it combines both technologies to get the best of both worlds today. 4GB of flash is cheap, and obviously, so are mechanical drives. Combine them to get similar performance to a SSD and reliability of an HDD.

What makes this special is not just that it has a cache. Every HDD out there has a cache. This puppy has a "cache" 100x what current drives have. What's more is that this cache is persistent/nonvolatile. It's good when you reboot, so even at OS load, you see the advantages.

Re:4GB?

[Bigjeff5]

Yeah, that wikipedia article is using figures from over a decade ago (1997 for their highest NAND numbers). Today the number of write cycles for the premium chips are in the millions, not thousands. You'd have to write your entire 500gb hard drive 8,000 times to burn up that 4gb of flash, assuming 1 million write cycles. In these applications, though, they're probably using the "creme de la creme" of flash chips, which can last up to 5 million write cycles. That lets you write that 500gb hard drive 45,000 times before the flash dies. That's 22.5 petabytes worth of data transfer. Assuming an 11ms write time for the hard drive (typical of notebooks), and assuming I didn't royally screw up somewhere (I may have), I get 32 million years of non-stop writing to use that up.

Even at their slowest, SSD drives are 5-10x faster than spinning disks. This allows it to act as a perfect layer of cache between RAM and HDD. It's also non-volatile, like the rest of the hard drive, so all you need to ensure no loss of data is to at least get the data to the cache. Assuming the hard drive doesn't do something retarded, like wipe the cache on reboot, it's all good.

Re:4GB?

[ooshna]

Pfft nerds reading books that are on paper yeah maybe 10 years ago. Your nerd card can be turned in at the front desk.

Re:4GB?

[i.of.the.storm]

Actually, since semiconductor manufacturing gets cheaper practically exponentially (yay Moore's law) eventually SSD prices should catch up and undercut mechanical disk prices, just because of the manufacturing process. But that's a long ways off yet.

Re:4GB?

[hairyfeet]

Actually that link just proves that Wikipedia data can be as old and out of date as any dead tree encyclopedia. But here is what I don't get: nearly all laptops nowadays have have built in card readers, yes? Why not simply get a fast 8Gb (or bigger if you find one on sale) card and use Readyboost for the caching?

I've found using W7 Perfmon that Readyboost tends to get around 2x on the cache thanks to the compression, so an 8Gb will give you roughly 16Gb of Readyboosted cache, and if you pound the cache who cares? It is easily replaceable. While I can't say how good it works on laptops I can say that having 8Gb of Readyboost (with compression 16Gb) with 8Gb of RAM does seem to give the OS better responsiveness, especially when doing heavy I/O tasks like gaming and video editing.

So I'd want to know how this drive will work WITHOUT the cache, or even if it will work at all, before I'd think about it. I know many folks treat laptops as disposable now, I have many drives a decade old still purring along and I don't know if I'd trust my data to a drive with SSD for buffering. Does the drive warn you when the cache gets below a certain percentage so you can replace it? With Readyboost I can scan the cache myself and if its size starts going down then I can replace it easily.

Re:4GB?

[eggnoglatte]

But they'd also have to use the SSD for read caching in order to get the benefits they advertise. And *reading* the contents of your hard drive 8000 times is maybe not that far off... Under most usage scenarios, the drive will still last 2-3 years, but it certainly has a more definitive expiry date than most other hard drive technologies (including pure SSD, which only degrades on write).

Re:4GB?

[WuphonsReach]

Actually, since semiconductor manufacturing gets cheaper practically exponentially (yay Moore's law) eventually SSD prices should catch up and undercut mechanical disk prices, just because of the manufacturing process. But that's a long ways off yet.

Well... maybe. NAND flash isn't too many generations away from hitting a lithography wall from what I remember. And we're still 10x-20x more expensive then magnetic media on a $/GB measurement. 2.5" laptop drives are in the 0.14-0.17 $/GB range, while SSDs are still up in the 1.36-2.34 $/GB range.

Assuming that we're already down in the 40-50nm(?) range for flash media, can they really manage to squeeze one more order of magnitude out of the feature sizes? (Closer to 3x smaller features since you double the bits at sqrt() feature size.) If they can make NAND flash down at 10-15nm, then yes we might see $/GB prices below $0.20.

25nm is apparently around the corner, so we might see a 2x-4x price improvement soon. The question is how much headroom is left after that.

Re:4GB?

[afidel]

No manufacturer of NAND cells claims millions of write cycles, zero. Even the best SLC designs are in the hundreds of thousands of cycles. You will also note from the benchmark numbers that the flash doesn't really help the random access time for this hybrid drive, ~60ms read and write at a paltry 100 IOPS, which is solidly in the slow SATA drive category.

Re:4GB?

[BikeHelmet]

What makes this special is not just that it has a cache. Every HDD out there has a cache. This puppy has a "cache" 100x what current drives have.

I think it's silly. The NAND will wear out really quick, and there doesn't seem to be much of a performance boost over a fast WD Black. Actually, there's no performance boost - but it is a smaller form factor that fits in laptops.

I remember a few years back a (Japanese?) company paired 1-2GB of cache with a 5400RPM HDD. They completely maxed out SATA1's transfer speeds - faster than the fastest WD Black released today. (The 2TB one)

If we had HDDs with 2GB of cache rather than 32-64MB, our drives would be a lot quicker. Quite possibly comparable to SSDs in sequential read/write, and a bit higher than our current HDDs in IOPS. I'd take one simply because it wouldn't wear out like NAND does.

Re:4GB?

[InfoJunkie777]

I read TFA. The tests were inconclusive. And the drives have to "learn" (most accessed data is written to NAND after a few calls). I like the hybrid concept and anything to boost performance on laptops drives is great. That being said, I would rather see the SSD used like ROM for the OS for FAST boot-up times. Surprised they didn't do it that way.

Re:4GB?

[i.of.the.storm]

Well, that is true. But depending on who you talk to, some people will tell you that it's not unlikely that we'll find a way around the wall when the time comes. I myself don't know much about process technology, but one of my electronics professors would occasionally mention that we've thought that there were dead ends in the technology before, until someone made a breakthrough discovery and found a way around the roadblocks. However, we are indeed getting to a point where we're limited by the size of the atoms themselves, which is problematic. But the future is exciting and unknown, so who knows what will happen next.

Re:4GB?

[LoRdTAW]

Sorry, its already been revoked.

Re:4GB?

[countach]

But what good is cache if it isn't persistent? The OS already has a perfectly fine read cache. It's the write cache that is the problem, and a non-persistent write cache of multi-gigabytes is pretty scary if you suddenly lose power. You could wipe out an entire file system that way.

Re:4GB?

[BikeHelmet]

But what good is cache if it isn't persistent?

Buy a UPS! ;)

And slap a huge warning sticker on the HDD!

Or maybe put a 9v battery in it, in case the power suddenly goes out?

There's plenty of ways to tackle the issue. How do you tackle silent data corruption? I've experienced it first hand running a linux distro off an SDHC card. It's easier to deal with power failure.

Re:4GB?

[Vlado]

From what I know the problem with SSD is only related to write-based access. So unlimited reads should not pose an issue.
Also, I think it's quite realistic to assume, that there's some kind of logic in place that will prevent data movements to bad blocks. In practice that would mean that while the 4GB of cache might "go bad" over a few years you would essentially end up with a standard 500GB disk. And, according to the tests, even on first access (before it learned where to put "busy" data) the disk performed a bit better that standard competition.

Re:4GB?

[Calinous]

Some SSD already have a supercapacitor inside them (just like a battery, hopefully without the sudden drop in usable capacity when you most need it)

Re:4GB?

[deroby]

The idea indeed isn't new and although I prefer eBoostr over ReadyBoost because it 'feels faster', the main hurdle on both technologies is 'outdated information'. The data on the (external) cache cannot really be trusted when you reboot (or re-insert the cache-device). I'm not sure how either technololgies overcome this, but having the cache sit 'closer' to the actual harddisk should make it a lot more reliable imho

Gets Better Over Time

[WrongSizeGlass]

The performance of the drive gets better over time as it 'learns' your most frequently used files. I hope it's smart enough to ignore the 'swapfile'.

Re:Gets Better Over Time

[yincrash]

Wouldn't you want your swap space to be stored on the faster SSD rather than the slower spinning media?

Re:Gets Better Over Time

[Anonymous Coward]

Only if you want the SSD to die within a month. Windows writes to the swapfile pretty much constantly.

Re:Gets Better Over Time

[eln]

These days with RAM being so cheap, your swap space is basically a warning that things are going terribly wrong. You want your swap on slow storage because slow storage is cheap and your swap should see very few writes under normal operation. If your machine starts hitting swap like crazy, you'll know immediately because your performance will go straight down the crapper as it feverishly tries to write to slow storage. This is your cue to figure out what's wrong and fix it ASAP so your machine will stop thrashing.

Re:Gets Better Over Time

[Anonymous Coward]

>Only if you want the SSD to die within a month. Windows writes to the swapfile pretty much constantly.

A month is probably exaggerating. I think Intel, for example says their mainstream drive should last at least 5 years with 20 GB:s written every day.

Re:Gets Better Over Time

[AvitarX]

Swap space can easily outpace that though.

Though I suspect default settings in operating systems will trend towards not doing such.

Re:Gets Better Over Time

[AusIV]

The main thing I use swap for these days is hibernating my laptop. What I need is persistent storage - the quicker the better.

Re:Gets Better Over Time

[N Monkey]

Only if you want the SSD to die within a month. Windows writes to the swapfile pretty much constantly.

Oh dear, I'd better remember to replace my SSD-based laptop, err, 9 months ago.

Re:Gets Better Over Time

[Rogerborg]

Ooh, good idea. By the way, I've been thinking of making an alarm clock that electrocutes your testicles if you hit the snooze button - can I sign you up for the beta?

Re:Gets Better Over Time

[Anonymous Coward]

Not neccessarily. I've got a bunch of drives in my computer, and each one is host to a small swap file. So if one link is satuarted for whatever reason in theory the computer should be able to write to the other swap files without any issues.

Re:Gets Better Over Time

[TheRaven64]

Easily? I think not.

Swap tends to be demand paging, where you're loading in a page as it's accessed (and writing a page out to make room, although this happens asynchronously with a little buffer). Satisfying page faults from a drive with a 5ms access time limits you to around an 800KB sustained transfer rate for reading. If you're lucky, you can do sequential writes so you keep your 800KB/s write rate. At this rate, you'd only use 7GB/day on swapping, if all of your disk I/O is swapping and you are swapping constantly for 24 hours. If you're writing more than a few hundred MBs per day to swap then you are going to notice that your machine has really terrible performance.

Re:Gets Better Over Time

[AvitarX]

God I hope to be modded down, thanks.

Re:Gets Better Over Time

[VanessaE]

You're forgetting one thing:

Sometimes, a machine will go from seemingly normal to suddenly thrashing about in swap rather heavily, with no warning at all. This has been the bulk of my experience, anyway. When your machine gets to that point, and you're in a graphical environment like the majority of desktop users are, you may not be *able* to look into the problem at all. You have to wait until after the damn thing comes to its senses, because you can't even switch to a regular text console, let alone log in in from another box. Forget trying to spawn a terminal. Every little program you launch to try to find the cause just causes the machine to use more memory or swap at this point, which just compounds the problem.

When the offending program finally does end, it's too late to see what went wrong because most programs leave no traces of their actions other than doing whatever they're programmed to do. Unless you're running some kind of process/resource logging program on your box (I'm not aware of anyone who does this outside of security professionals perhaps), good luck finding out what actually caused the problem, unless you saw something visibly bug out just before the machine stopped responding.

There are no two ways about it - this is absolutely the worst way to handle an out-of-memory condition. Most people would much rather have programs complain about lack of memory than to have their machine "lock up" for an hour while sitting there churning away in swap. In my experience, the average user figures their computer's being stupid again, and it's time to hit the power switch or the reset button, or maybe call someone for help (which doesn't work anyway, so they're back to square one).

To give an example, I once set my machine off to run a build which should have taken maybe half an hour, and went off to run some errands and watch a movie. It was still going three hours later, and had dug the machine so deep into swap that mouse events were taking 10-20 seconds just to echo to the screen, and keyboard events were nonexistent as far as X was concerned. I tried my level best to bring the machine back to a sane state, but I eventually had to give up and hit it with Alt-SysRq-U/S/B.

I love Linux as much as anyone, but I got sick and tired of this happening on my boxes, and responded the only way that seemed to make sense: I disabled swap entirely on both systems and added enough RAM to each to make up for the lost "memory".

Aside from older hardware that clearly needs it because of sheer lack of RAM, is there even any reason to recommend/enable swap by default anymore? Modern machines come standard with around 4GB of insanely fast RAM - isn't that enough?

Re:Gets Better Over Time

[BikeHelmet]

From what I've seen on the NCIX forums, heavy use seems to kill the Indilinx ones in about 15-20 months. A lot of people have had to RMA them...

Re:Gets Better Over Time

[Hurricane78]

No, it gets worse. Because SSDs are known to massively lose performance over time. Down to 1/8. Yes, even the newest ones.
And then they fail. Badly.

I won’t be touching an SSD, until it can *guarantee* me equal or better reliability than hard discs before those also started to get worse (some years ago). My data is too valuable for experiments. Sorry.

Re:Gets Better Over Time

[Hurricane78]

Good luck compiling GHC then. Or running a couple of big Java programs. Or both. Plus a virtual machine.

I have seen RAM usage go up to 8GB and beyond. I don’t have 8GB. So I’m happy to still be able to do it at all.

Re:Gets Better Over Time

[Anonymous Coward]

read up on the 'limit' command... set memoryusee/vmemoryuse etc. so something other than unlimited... then processes will get killed if they grow so big as to push your machine into swap death

Re:Gets Better Over Time

[Anonymous Coward]

[quote]Sometimes, a machine will go from seemingly normal to suddenly thrashing about in swap rather heavily, with no warning at all. This has been the bulk of my experience, anyway. When your machine gets to that point, and you're in a graphical environment like the majority of desktop users are, you may not be *able* to look into the problem at all. You have to wait until after the damn thing comes to its senses, because you can't even switch to a regular text console, let alone log in in from another box. Forget trying to spawn a terminal. Every little program you launch to try to find the cause just causes the machine to use more memory or swap at this point, which just compounds the problem.[/quote]

DTrace. This is exactly the kind of situation it was engineered to cope with.

Re:Gets Better Over Time

[izomiac]

Umm, I've have a few harddrives die on me. They're probably the least reliable thing I own. When a HDD dies you've pretty much lost everything on it, and I certainly wouldn't trust one past a few years... So, why do you expect an SSD to be worse? Electronic failures can happen on either (should be identical risk), but there's no wear and tear on an SSD. If you reach the 100,000 write cycle limit (most pessimistic estimate I've ever seen) on a (tiny) 32 GB drive after a year of writing at maximum speed (?!?!?), then it's read-only. Performance loss IMHO, it sounds like you're suffering from the zero-risk bias, or you're not migrating for other (probably valid) reasons.

Performance-wise, that could happen. You can fix it easily enough with SATA secure delete or TRIM (HDDs aren't maintenance-free either, hence why they need defragmentation). It doesn't always happen, my two-year-old $50 30 GB SSD has been through a couple dozen OS installations and a fair amount of SVN downloading of large projects, and it's still as fast as advertised (I just double checked).

But, more importantly, WTF are you doing not backing up your valuable data?!?!?! Drive failures are all but inevitable with conventional HDDs, and I'm certainly not even going to risk my data on SSDs being infallible.

Re:Gets Better Over Time

[petermgreen]

These days with RAM being so cheap
Ram is cheaper than it used to be years ago but:

1: In the best case ( DDR2 or DDR3 desktop modules) it's still about ten times more expensive per gigabyte than SSD.
2: Often the systems you have aren't the best case. Older motherboards often have hard ram limits that are lower than you would like and/or take older ram families that have a much higher cost per gigabyte than current ones. This can be alleviated by a new motherboard but that may in turn mean a new case, possiblly a new copy of windows and possiblly some other devices replaced as well. Laptop users have it even worse. You can't go above 8GB on any laptop i've seen and even 8GB is both expensive and only available on modern systems.
3: A lot of people are still stuck with the limitations of 32-bit desktop windows (For example the new data aquisition module I recently bought is refusing to work on my 64-bit windows install despite supposedly having 64-bit drivers)

Re:Gets Better Over Time

[rdnetto]

Aside from older hardware that clearly needs it because of sheer lack of RAM, is there even any reason to recommend/enable swap by default anymore? Modern machines come standard with around 4GB of insanely fast RAM - isn't that enough?

I don't know about other OSes, but Ubuntu needs it for hibernation (and requires that it be a partition, not a file). Complete waste of 4 GB if you ask me.

Re:Gets Better Over Time

[RichiH]

By "complain" you mean "get hit by the OOM killer", of course.

I would rather ctrl-alt-f1 to a tty and fix things than have them dying randomly.

Re:Gets Better Over Time

[Anonymous Coward]

Only if you want the SSD to die within a month. Windows writes to the swapfile pretty much constantly.

Oh dear, I'd better remember to replace my SSD-based laptop, err, 9 months ago.

Windows comes for FREE with Balmer GREMLINS which automatically and very efficiently DESTROY YOUR HARD DRIVE !!!!!!

Re:Gets Better Over Time

[maraist]

top
M
k
copy-paste ID

works pretty well for me in GUI's

I could figure out the killall / ps equivalent, but I'm too lazy.

Course I always have 30 terminal windows already open. Can't open a new one in swap-hell.

Re:Gets Better Over Time

[Anonymous Coward]

Games that pre-load 10GB of data into memory and manage swap access at the same time.

Jimungous excel files accountants want to open and edit on the fly without spending $$$ on hardware.

Databases where the whole shebang can't be loaded into memory and its a whole lot faster to place semi-used-a-lot stuff in a place you know where it is e.g. swap.

Instances where you KNOW you're asking your computer to do something outside of the realm of normality and you don't want it to crash with a memory error.

As memory has gotten cheaper and faster the disparity in speed between memory and hdd's has gotten wider; the fastest disks have gone from ATA PIO mode 1 @ 3.3mbps with (realistically) 200 mbps throughput memory (PC33 ) (60 to 1) to 120mbps (max, I'm being nice) disks with DDR3 12800MBPS throughput (106.6 to 1). SSD's help but you need RAID SSD's to get back to a reasonable throughput ratio. Speed, not size, IMO, is the real bugger here. Plus latency for random reads/writes hasn't gotten much better since those old drives. Still takes awhile to go from one end of a platter to another.

Try adjusting the swapPINESS

[tepples]

What swapfile? I have used Ubuntu on a few PCs with at least half a GB of RAM, and I rarely see swap usage climb above 40 MB. In an environment where reads are cheaper than writes, you'll want to use a low value for the swappiness, such as 10% instead of the default 60%.

Re:Try adjusting the swapPINESS

[BrentH]

Now I know it said swap just before 'pines', but thats still not 'penis' yo.

Re:Try adjusting the swapPINESS

[MrNemesis]

The problem with ubuntu (indeed, most Linux distros I've used) is that they also sue the swap file as the hibernation file - last time I set up a laptop, I wasn't able to hibernate due to my habit of capping swap partitions/files at 256MB*, and last I looked it was impossible to set it up to use a swapfile due to the way suspend was implemented, it had to be a partition. More confusingly, linux (judging by the amount that's written to disc when hibernating) writes the entire contents of memory, rather than dumping buffers+cache as windows seems to do, but that's another story. I've not had too much time to play with various suspend solutions but that's how it seems anyway.

I know loads of people will chime in with "disc space is cheap! Why not have an XGB Yfile?!", but when you're using SSD's, it's not cheap or plentiful.

What I'd love is the ability to have a minimal swap partition and a dynamically resizeable hibernate file that I can delete and have automatically recreated if need be - heck, the default 5% root reservation used by ext filesystems should ensure there's space available on the drive if need be.

* To everyone that thinks you should have 2*$yourmem in swap, IME you're almost certainly wrong on anything with 2GB or more of RAM (my bloaty-as-hell KDE4 workstation with 8GB uses about 1.5GB as a working load and my debian backup box only uses about 75MB of RAM for actual programs, the rest is all used as cache). The only time I ever have apps that run out of memory is when one develops a leak or somesuch (hey! Thanks for eating up all my 16GB of memory and 32GB of swap, BackupExec!), in which case I want the OOM Killer to terminate it ASAP, rather than have the discs grind away for an hour first, making the computer more or less unusable. If anyone can find me a situation where having gobs of swap is handy I'd love to hear it...!

Re:Try adjusting the swapPINESS

[Striek]

While this theory holds for a single workstation, it doesn't work so much when you throw VMs into the mix. I have physical servers hosting dozens of virtual machines. This means I can overcommit RAM for two reasons:

1. Some machines are used infrequently, and are better off sitting in swap, leaving RAM space available for more frequently used machines.
2. Several of them quite often run identical operating systems, and memory pages can be shared between them with an appropriately configured hypervisor.

Whereas before, it was inactive applications that got swapped out to disk, I'm finding more and more these days that it's inactive servers that get swapped out. Storing the swap on SSD's has significantly reduced the time to live on these machines. And while a few gigs of RAM is cheap for a workstation, 64 Gb of fault-tolerant server-class RAM... not so much.

Using SSD's to house swap memory, I've found I can maintain the same level of performance while requiring less RAM.

Re:Try adjusting the swapPINESS

[the_one(2)]

It's probably not hard in the least to create a swapfile on hibernate and then delete it on resume with very simple scripts. I could be wrong though. Reserving space might be harder.

Re:Try adjusting the swapPINESS

[tepples]

Whereas before, it was inactive applications that got swapped out to disk, I'm finding more and more these days that it's inactive servers that get swapped out.

Incidentally, I've seen this happen on a Go Daddy virtual dedicated server.

Using SSD's to house swap memory, I've found I can maintain the same level of performance while requiring less RAM.

How fast do they wear out?

Re:Try adjusting the swapPINESS

[Hurricane78]

You just don’t really use your PC. That’s all. ^^

Real men always go over the limit. Yeah!

And I can easily fill over 2GB in default operation, without even anything unusual running.
That’s because my motto is, that if I have it, then I’ll also use it! :)

Re:Try adjusting the swapPINESS

[Striek]

How fast do they wear out?

I've only been using SSD solutions for about a year now, so there's not much of a dataset yet. But I'd fathom a guess of about 3 times faster than mechanical hard drives currently - worth it, I think, for 100x the read speed. Last year when I looked at the research, the numbers all said 10 times faster. It's definitely improving.

Manageable hybrid

[Thanshin]

Hybrid storage drives should be manually manageable.

You should have the possibility of configuring which files/folders/partitions/whatever you want to be accessed fast and which parts are to be left as "long term", slow access, storage.

Re:Manageable hybrid

[Anonymous Coward]

I have a manageable hybrid.

Read heavy system partitions on a small SSD (/boot, /bin, /etc ...etc), everything on magnetic.

Re:Manageable hybrid

[ThoughtMonster]

Not feasible on a laptop. This drive, however, seems perfect for most purposes, and the price is not half-bad either ($156 for the 500GB version). Performance is better than traditional drives in almost all benchmarks, and reaches up to 2x the performance when using commonly accessed files (like, the operating system).

I'd love to see what this does for boot speeds etc.

Re:Manageable hybrid

[Thanshin]

I have a manageable hybrid.

Read heavy system partitions on a small SSD (/boot, /bin, /etc ...etc), everything on magnetic.

Separating both parts leaves you with two options.

A manageable hybrid would let you have more degrees of speed/size. It would let you use just a part of the SSD to accelerate the accesses to a part of the magnetic storage, the rest as pure SSD and the rest of magnetic for low priority storage.

Re:Manageable hybrid

[bluesatin]

So you mean like having 2 separate drives?

Re:Manageable hybrid

[Thanshin]

So you mean like having 2 separate drives?

Two drives with a direct connection that allows me to seamlessly save what I'm reading on the SSD so a second access is faster.

(while having all other options of having them both)

Re:Manageable hybrid

[Alien1024]

That's probably not a hybrid but two drives: a SSD and a magnetic drive.

Re:Manageable hybrid

[dingen]

Except it would not be two physical drives. Awesome for laptops!

Re:Manageable hybrid

[drerwk]

I feel the same way about swap pages.

Re:Manageable hybrid

[boxwood]

hard to fit 2 drives into a laptop.

Re:Manageable hybrid

[Anonymous Coward]

thats the old idea of swapping. whats used now is rather called paging. meaning any new program starting up in memory will equally reserve space on swap. thus the constant need to access swap space.

Re:Manageable hybrid

[Anonymous Coward]

For this to work, the drive would have to have a clue about filesystems which they don't. Anything of this nature has to be done in the OS, not the drive

Re:Manageable hybrid

[Kjella]

If you mean that applications or the OS should be able to give cache hints, then I agree. If you want essentially two drives and a million manually managed symlinks I don't want that, it'll be a 10% overhead managing to save 1% on performance. And if you're only doing coarse level like installing one app here and saving files there then get two disks (there are dual bay laptops too).

Re:Manageable hybrid

[VinylPusher]

Depends on the laptop.

I have 2 in mine. Others are able to support 3.

Re:Manageable hybrid

[ZorbaTHut]

I sort of disagree. Humans are really, really bad at this kind of management, and a smart computer algorithm can often do better. Just look at the people who disable swap space because "it makes the computer slower". You can't trust humans to manage this optimally, and computers can, in theory at least, generate extremely complicated structures and processes (i.e. "if the user runs this program, he's probably about to be reading this data, so let's get this onto the SSD ASAP.")

Re:Manageable hybrid

[TheLink]

I actually wondered if unionfs would support it well- combine SSD with HDD.

As for the OS giving the cache hints, it would be nice if the OS would do that but often the OS doesn't know either - it is also guessing what the Applications and Users are up to.

If a hybrid drive was aware of popular filesystems, there's a fair bit it can do to make things faster, especially if it was for "single user desktop use".
1) You'd have a small nonvolatile write buffer that'll cache all writes, so that you can more easily figure out whether stuff is sequential or not. Only write to disk when buffer is full.
2) You'd not cache sequential writes if you just wrote something that looked like a new big/huge file was being created (from the metadata) - those writes go to disk.
3) You'd try to figure out which sequences of block/file accesses are popular and have sequences of blocks that are fragmented - cache the blocks that would cause big seeks, then cache the entire sequence of blocks if they turn out to be really popular.
4) You might want to cache file system metadata and stuff like the registry (which tends to be fragmented) - this helps when you haven't figured out 3) yet.

Decades ago I wrote disk caching software for the Apple IIGS that just cached metadata and the current track the head had read. It actually was aware of the file system format and conventions. It worked reasonably well for the time - files were smaller back then so less fragmented, the big slow seeks were mainly due to reading metadata. Also on the Apple IIGS a lot of the drive reading was done by the CPU+OS. So if you were looking for one sector in a track, you'd have to move the drive head to the track and keep reading till you get to the sector you want. The cache software I wrote would also cache the "unwanted sectors" that went past the head - might as well since the IIGS had a lot more free RAM than the older Apple IIs (a 48K RAM or smaller computer does not have spare RAM to "waste" on a track cache). Back then there was stuff like sector interleaving because some OSes were too slow to keep up with processing the sectors, so if you wanted sector 1 then 2, having them consecutive would mean you'd have to wait for the disk to spin around again... Track caching reduces the impact of that.

Re:Manageable hybrid

[petermgreen]

Just look at the people who disable swap space because "it makes the computer slower".
There are two main mindsets to designing computer systems.

The batch processing mindset says that what matters is average performance.
The real time systems mindset says that what matters is meeting your deadlines consistently.

IMO desktops are closer to the latter than the former. Tens of milliseconds on each user action won't generally be noticed, the user can't do the next operation that quickly anyway. Tens of seconds on one action WILL be noticed and quite possiblly piss the user off especially if it's unexpected even if it only happens on a very small subset of actions. Unexpected delays break the flow of thought.

Now consider an app like firefox. It has a habbit of using a LOT of memory (whether this is a leak or a design feature is a subject of many /. arguments and not one I want to get into here). It is also single threaded so if any part of the app needs something swapped in the whole app is blocked. If the OS decides to swap it out for whatever reason (e.g. some app ran away with memory usage and didn't finally fail until after it had swapped out everything or a long running batch job overnight caused the OS to swap stuff out and expand the disk cache). Then you click on it's taskbar icon and wait ages as all the memory pages it's state is spread over grind their way back into memory.

You can't trust humans to manage this optimally
True but you can't really trust computers to either. Especially when the computer hasn't really been told what the human considers important or even how the data will be used.

Re:Manageable hybrid

[rastoboy29]

You mean, like separate drives?

I, too, question the real value of this technology--but what you suggest would defeat any conceivable purpose for it.

Hmmm...

[Anonymous Coward]

Wouldn't it be cheaper and more effective just add an additional 4 GB of memory for disk cache?

Re:Hmmm...

[Lennie]

It wouldn't help start up time would it ?

Re:Hmmm...

[Anonymous Coward]

Neither will an in-disk cache. You access the start-up files only once per boot, not a likely cache candidate.

But to the OP - with 4GB on flash, you boot up with a warm cache (for your regularly used data/programs), where as the RAM cache will need time to warm.

Re:Hmmm...

[dingen]

No, because then every time your computer reboots, you need to fill the cache again. Using a solid state cache, you need to fill the cache only once and then keep it updated according to your usage, but a reboot wont harm performance at all.

Re:Hmmm...

[Yvan256]

Most laptops, netbooks and all Macs except the Mac Pro have very limited RAM upgrade capabilities. You can't "just add an additional 4 GiB of memory".

Re:Hmmm...

[Anonymous Coward]

He didn't say add 4GB of RAM. He said 4GB of hard drive cache. There is a huuuuuuuuge difference.

Re:Hmmm...

[InfiniteWisdom]

How would it be "cheaper"? 4GB of RAM costs about $100. That gets you about a 32GB SSD.

Several of the other comments below address the "effective" part.

Re:Hmmm...

[Anonymous Coward]

What is this "reboot" you speak of

Re:Hmmm...

[Yvan256]

How about adding both 4 GiB of Flash for those start-up times (good for laptops) and 4 GiB RAM for the hard drive cache? That would scream to say the least. However at that point I don't know if it would be even more expensive and less reliable (controller for efficient use of three types of storage?) than simply going with 100% SSD.

Re:Hmmm...

[Kozz]

No, because then every time your computer reboots, you need to fill the cache again. Using a solid state cache, you need to fill the cache only once and then keep it updated according to your usage, but a reboot wont harm performance at all.

Unless of course you spend most of your time on a work laptop that does not belong to you. Then after the initial boot, you click the shortcuts for Outlook, Firefox and Eclipse, and then get a coffee and work in your Sudoku book for 5 minutes while waiting for the gorram networking and policy-related scripts to finish... whatever the hell they're doing so your apps can actually start and you can begin working. Not that I'm irritated or anything.

Re:Hmmm...

[Anonymous Coward]

Yea.. that's just great.. Now it makes it even harder to diagnose system issue problems. Now we can possibly corrupted files that are now sitting in this hard to acces 'cache' that will survive system reboots.

Re:Hmmm...

[Rockoon]

Cheaper? Are you kidding?

I am amazed that people complain about an SSD's cost per gigabyte, and then haul off and say that memory is cheap.

Re:Hmmm...

[petermgreen]

He didn't say add 4GB of RAM. He said 4GB of hard drive cache.

He said "Wouldn't it be cheaper and more effective just add an additional 4 GB of memory for disk cache?"

Now the problem is how to add that memory? I see a few possibilities.

1: add it as extra main memory and let the OS use it for cache. The trouble with this approach is twofold. Firstly a lot of users are still tied to 32-bit windows and desktop editions of 32-bit windows cap out at 4GB of address space (which usually translates to somewhere arround 3.5GB of usable ram). The second is a lot of machines especially laptops can't take the extra ram and even where they can a pair (most laptops will already have both slots filled and want dual channel for best performance so replacing both modules is the only reasonable option) of 4GB laptop modules isnt' cheap (crucial says around $400 for DDR2 and $500 for DDR3)

2: add it to the storage controller. This is basically what high end raid controllers do. Not an option for laptops unless the laptop maker puts it on the motherboard. (which they are unlikely to do because of space constraints). For desktops you could drop in a raid controller that can take 4GB of cache ram but the price is insane (you'd probablly be better off just buying a 160GB SSD and putting your OS partition on it)

3: make it part of the drive. This is essentially what seagate have done here though they have chosen to use flash rather than ram (probablly because of it's non-volatility, higher density and lower cost per gigabyte).
4: add it as an extra device. Again unsuitable for laptops and likely to be a pain for desktops too. I'm not aware of any products offering this but if there were I'd expect them to be expensive.

ReadyBoost in hw?

[W2k]

I wonder if this is simply a more expensive version of ReadyBoost. Similarly, it takes your most frequently used files and puts them on a flash drive for faster access times, in a way that is transparent to the end user. In this case I wonder if there would be any speed gain from using this on a PC running Windows 7 with ReadyBoost? Caching always introduces some overhead, so rather than using multiple levels of "flash cache" it might be better to simply turn ReadyBoost off in that case. My experience with ReadyBoost has been that it does indeed improve performance, but in no way close to using a real SSD as the system drive.

Re:ReadyBoost in hw?

[eqisow]

Hmm, USB vs SATA... I imagine this would be faster than Ready Boost. On the other hand, if you're rocking USB 3.0...

Re:ReadyBoost in hw?

[Sockatume]

Microsoft actually did pitch "ReadyDrive" hybrid SSDs as a selling point for Vista back when it launched. It was basically the same as this, except the caching was controlled in the OS and not the drive and it did some fancier stuff like caching boot data on shutdown. It didn't do very well, perhaps because the technology wasn't mature enough in price and speed.

Re:ReadyBoost in hw?

[dingen]

Not to mention it would be a Windows-only product.

Re:ReadyBoost in hw?

[Yamata no Orochi]

I'm not sure why Microsoft producing a Windows-only product would in any way harm their sales.

Re:ReadyBoost in hw?

[dingen]

If Microsoft would produce them by themselves, that wouldn't be a problem. But as they don't do that sort of thing, they need to convince a harddrive manufacturer to produce it for them. And I don't think a lot of harddrive manufacturers would be interested in such a product.

Re:ReadyBoost in hw?

[Anonymous Coward]

I'm not sure why Microsoft producing a Windows-only product would in any way harm their sales.

I'm not sure why anyone other than Microsoft would view their sales figures as relevant to this discussion about interoperability.

Re:ReadyBoost in hw?

[Yamata no Orochi]

The micro-discussion was actually regarding Microsoft's ReadyDrive, and the mention that it was a selling point to Vista.

Actually, this whole thread is an offshoot of a comment comparing the drive's function to ReadyBoost.

Re:ReadyBoost in hw?

[dskzero]

Not to mention it would be a Windows-only product.

... and the point being?

Re:ReadyBoost in hw?

[dingen]

That Seagate's concept of a hybrid drive is a lot better product than what Microsoft was suggesting with the ReadyDrive concept, as Seagate's hybrid drive doesn't rely on functionality only provided by Microsoft Windows and Microsoft's ReadyDrive does.

Re:ReadyBoost in hw?

[Sockatume]

Well, precisely. A Vista-only product at that.

Re:ReadyBoost in hw?

[Yamata no Orochi]

You don't suppose Windows has enough PC marketshare to warrant producing hard-drives specifically compatible with a feature therein?

There sure were a lot of hardware manufacturers trying to fit inside that "Vista Ready"/Vista Compliant or whatever bullshit for that to be true, weren't there?

Re:ReadyBoost in hw?

[dingen]

Putting a sticker on your product is one thing, but I seriously doubt a lot of harddrive manufacturers would be interested in producing a product that would only function in Microsoft Windows. Especially since it would only work in Vista and 7, which together barely hold a quarter of the market today.

Re:ReadyBoost in hw?

[dskzero]

I do know that, and I do agree with you, but that was awfully off topic.

Re:ReadyBoost in hw?

[Bengie]

Since Vista already supported hybrid drives, I would *assume* the Windows would check for this stuff. Also, Vista/7 already disables readyboost automatically against drives that are SSDs.

Another issue that would crop up is SuperFetch runs under the ReadyBoost service and Superfetch is really nice.

Wikipedia: SuperFetch is a technology that pre-loads commonly used applications into memory to reduce their load times.

It actually caches commonly used files, not just "apps"

Essentially, it uses free memory to speed up your system by caching data directly into your RAM, which is even faster than any SSD. Sysinternals now has a utility that will show which files and offsets of those files are in memory. I noticed it has a ton of my game files loaded up in there which explains why my HD rarely reads while playing games.

Re:ReadyBoost in hw?

[MobyDisk]

Readyboost really doesn't help unless you have a really low-memory situation.

ReadyBoost goes through USB, which is a huge bottleneck. On top of that, most USB flash drives have really slow write speeds. ReadyBoost also has to read the file, then write it to the USB drive. Windows also must assume that the user could remove the flash drive at any point, so it can't cache writes. Integrating the cache into the drive solves all of these problems.

Re:ReadyBoost in hw?

[afidel]

You're looking at the wrong market, the correct market is what percentage of new HDD's are being sold for systems running Windows 7, the answer is ~90%.

Re:ReadyBoost in hw?

[petermgreen]

Nobody really knows what proportion of new machines are bought to run win7 because whether a machine is sold with XP or 7 and whether the customer immediately wipes it to load a corp image of XP or not it counts as a win7 OEM sale.

So they make a hard drive with a cache?

[Anonymous Coward]

And I am supposed to be impressed?

Can this HD do anything different/better than a HD with a dram cache?

Re:So they make a hard drive with a cache?

[dingen]

The difference between this drive and every other drive on the markt is not the fact it has a cache. Every harddisk has a cache. The difference is: this drive has a 4 GB cache. And because it's an SSD cache and not (as you suggest) a RAM cache, it maintains it's state even between reboots, so your computer is fast right from the start.

Re:So they make a hard drive with a cache?

[Anonymous Coward]

The difference between this drive and every other drive on the markt is not the fact it has a cache. Every harddisk has a cache. The difference is: this drive has a 4 GB cache. And because it's an SSD cache and not (as you suggest) a RAM cache, it maintains it's state even between reboots, so your computer is fast right from the start.

...now my computer is fast! finally fast!

Re:So they make a hard drive with a cache?

[Anonymous Coward]

I predict, initially, there will be an exploit related to this infrastructure change that will not be close-able with out negating the benefits therein.

Re:So they make a hard drive with a cache?

[Anonymous Coward]

2nd difference is that normal HDD has read cache only (obviously, power loss would demage files in cache). This disk has cache is for both read and write operations.

Or wait..

[XMode]

OCZ and im sure others have SSDs up to 500GB now. OK, they cost as much as my car, but they exist. It wont be long before they get up to 1TB, then 2TB.. Then its just a matter of waiting for the price to come down.

SSDs have caught up to traditional drives capacity extremely quickly, it wont be long before you can put a 10TB SSD in your laptop and never have to worry again (well, except for loosing it).

Re:Or wait..

[dingen]

SSD wont be as cheap per GB as traditional drives for many years to come. Chances are that even when a 500 GB SSD drive gets to an acceptable price point, an old-fashioned hard drive would still be cheaper and hold many, many more data at the same time.

This solution provides a cost-effective way to have both performance and storage *right now*.

Re:Or wait..

[inKubus]

it wont be long before you can put a 10TB SSD in your laptop and never have to worry again (well, except for loosing it)

Why are you worried about loosing it? Is there something on there the world should be afraid of? Should we all live in fear of the data to be unleashed? Are we all going to be affected when you let it loose?

Re:Or wait..

[dingen]

Should it be "much" instead? I'm not a native English speaker, so if you could enlighten me on how to use the language correctly, I would be thankful.

Re:Or wait..

[inKubus]

Multi-tier storage is all the rage in enterprise storage now. With a SAN (storage area network), it's fairly easy to have different types of storage pools available to all of your servers. Then it's just software performing a maintenance process to move data from front line to near-line to back line.

Typically this would probably go something like this, for a midsize organization with a few hundred TB of live data:

8-16 SSD's in a RAID10 (~500-1000GB)
32-128TB 15K SAS Drives RAID10
32-128TB 7.2K nearline SAS Drives RAID 6 or 10
32-??TB Tape such as LTO

There are completely autonomous systems that do this now. It's pretty complex to set up but it's possible now to get a multi-tier data setup for not that much money.

What this drive does is bring some of that power to the laptop world. If you have a desktop and haven't gotten a 40GB SSD yet, what are you waiting for? Granted, these Intel drives are not the fastest, but 40GB should be enough for almost anything you could do. You could definitely install an entire distro pretty easily. After that, if you run out of space, if you're using LVM you can just add another SSD that's twice as big in 2 years for another $100, then keep leapfrogging for $50/year to unlimited storage. Use some type of network storage if you have a ton of extra data, or add a 5400 rpm "green" drive.

This would work for most people and I predict 90% of desktops will be SSD-only in about 2 years. That's why hard drive manufacturers are scrambling to release drives in the 2TB, even 3TB range now. Their days are numbered. Cool, power efficient, fast SSDs are simply superior for almost everything and frankly are of a size where they are useful and competitive with any spinning disk for most desktop applications. Now, people with huge data like video and audio and the like may still need a drive or two in the home (or on the Internet somewhere, if they have a fast enough network), but the days of huge, mostly empty hard drives spinning 24/7 are coming to an end.

This might change culture a little bit. I do like the ability to hold a lot of data myself, but most people don't want to bother with it. This brings up a lot of obvious concerns about privacy, and that's going to be a war the drive manufacturers will be fighting with their own enterprise arms who want to sell drives to the data centers and cloud services that will be providing storage to the desktop users. Of course, the data won't be "yours" any more, it'll be a licensed copy streamed to your computer over iSCSI type connection.. That's what they all want now: Google, Microsoft, Amazon, Apple, Time Warner, Comcast, etc.

Re:Or wait..

[bblount]

Yeah, much is correct. "... would still be cheaper and hold much, much more data at the same time."

Re:Or wait..

[dingen]

How can I tell when to use "many" and when to use "much"?

Re:Or wait..

[Anonymous Coward]

Technically, data is plural (of datum), so "many" is correct. In practice it just sounds weird though, so everyone uses "much".

Re:Or wait..

[Jeff DeMaagd]

and...?

I don't see how the existence of large SSD invalidates the hybrid drive. What can invalidate the hybrid drive is if it does not deliver on its performance claims. Previous incarnations didn't, but maybe Seagate has something. If they can deliver a good percentage speed increase that beats the increase in cost, then they have something. Most files don't need SSD-like speeds, if the commonly used boot and application files are on SSD portion, then that would provide most of what I want on SSD without requiring that I buy a $1000 drive to get an SSD that meets my capacity needs when a $120 could have done the job almost as well.

A 10TB portable SSD is roughly five years away yet. We'll see by then whether the costs are competitive or not. We'll also see then how much data storage needs scale up then.

Re:Or wait..

[Arthur Grumbine]

Should it be "much" instead? I'm not a native English speaker, so if you could enlighten me on how to use the language correctly, I would be thankful.

"Much" would be correct. In general, "much" is used when speaking without reference to any kind of unit (e.g. oil, power, land, money, data, etc), while "many" is used when speaking with units (e.g. barrels, kilowatts, hectares, dollars, bytes).

Re:Or wait..

[dingen]

Nothing sound weird to me, as I'm not a native speaker. I understand data is plural, but do you also use it like that in sentences? Would you say: "the data is stored safely" or "the data are stored safely?"

Re:Or wait..

[dingen]

Thanks! So it's "much data" holding "many bytes", right?

Re:Or wait..

[NatasRevol]

When the grammar nazis attack you on Slashdot...

Re:Or wait..

[arielCo]

Long ago "data" was just the plural of "datum" (single fact or piece of information), and in that sense "many" would be adequate. But when using it as a mass noun (as is the case here) it would be a mass noun (measureable and singular like "water"), and you'd talk about "much data".

Re:Or wait..

[Anonymous Coward]

My cup holds much more water than your cup. My cup holds many more pebbles than your cup. I suppose "data" is more like water, in that there's a whole shitload of it and functions like a singular noun. Theoretically, though, data is a plural of datum. I bet back in the day when 4k was a lot, "many data" was a valid term.

Re:Or wait..

[Spatial]

"Many more data" is technically correct, but in casual conversation most people treat it as a mass noun rather than a plural of 'datum'.

Re:Or wait..

[Anonymous Coward]

You use "many" for countable nouns, and "much" for things you cant count.

example:

Countable nouns: Apples, dogs, cars. You can say "I have three dogs."

Un-Countable nouns: Sugar, money, furniture. You cant say "I have three sugars". You'd have to put something countable in there, like "I have three cups of sugar." to count it properly.

How "many" dogs? Three.

How "much" sugar? A lot.

So in the example you initially used, "data" isn't countable. You cant have "three data". So it would be "much more data".

If you changed it a little, you could say "three bits of data", because you can count a bit. Then it would be "many more bits of data". Though, no one really says it this way.

http://www.learn-english-online.org/Lesson36/Lesson36.htm

Re:Or wait..

[twidarkling]

A guideline that will serve well (but not perfectly) would be "many" for countable things, "much" for more ineffable concepts. You have many rocks, many people, many grains of sand, but much stress, much water, or much evidence, since you can count individual rocks, people, and grains of sand, but you can't count stress, water, or evidence (though you can count *pieces* of evidence, thus why this rule isn't perfect. In "pieces of evidence," you're talking about "pieces," and "of evidence" is modifying that, so you're using "many" to talk about "pieces.") I hope that helps somewhat.

Re:Or wait..

[twidarkling]

English treats this strangely. Either could technically be correct, since it depends on whether you're talking about the data as a single unit, or as a collection of individual datums. If you're talking about it acting as a single group, then you use the singular "is". If you're talking about it as a group of individuals, then the plural "are". Basically, if you were talking about several different groups of data being stored safely, say, your photos, your company's financial records, and your insurance, that could be a case for "the data are stored safely," especially since they'd likely be stored separately, in different fashions.

Re:Or wait..

[Bigjeff5]

It would have been correct if he had said "many more datum", but that's just weird.

Yes, "much more data" is the correct way to say it, which was the point of the GP's joke.

Re:Or wait..

[Bigjeff5]

Many is for individual items, much is for groups of items or non-singular items. Data is a word that means many pieces of datum. It's not a plural of datum in the traditional sense - you treat it like a new word that means "many datums" instead of a plural form of datum.

I.e. - I have many knives in my drawer, but there is much tea in my glass, and not much time to drink it. The knives are referred to as individuals, even though there are many of them. The tea is a liquid, and is composed of individual molecules, but we don't count them, we refer to them all as a group - tea. Time is an idea that cannot be singular, so it is treated the same as a group.

The real trick to English, though, is that for every rule, there is an exception (even this one). Most native English speakers don't speak the language correctly anyway, so don't worry too much if you mess up something like using "many" instead of "much".

Re:Or wait..

[Try_Nice]

How can I tell when to use "many" and when to use "much"?

Like the others have said, it has to do with whether something can be counted or not. Conversationally speaking, data is not countable, because no one ever uses the word "datum" in casual conversation. (You are among a very few English speakers that even realize that data is plural and that datum is the singular!)

Many more gigabytes / much more data.

A cool tip! As confusing as the words "many" and "much" are, you can safely just use "a lot" instead as it always works. A lot more gigabytes / a lot more data. :-)

Re:Or wait..

[dingen]

Thanks, I suppose I'll be using that a lot from now on.

Re:Or wait..

[Bigjeff5]

Data is a sort of "group" word - the word is singular, but it stands for many individual pieces of datum.

You treat it as a singular word in every case, therefore "data is stored safely". People who say "data are ..." are incorrect, unless "data" is just being used to describe something else. I.e. - "Several chunks of data are missing from the database." The relevant phrase there is "chunks are missing". "Of data" just describes the type of chunk.

Re:Or wait..

[Bigjeff5]

I imagine an SSD drive much like Iron Man's suitcase, except when it unfolds it turns into a giant, Tokyo destroying monster!

Re:Or wait..

[BikeHelmet]

This solution provides a cost-effective way to have both performance and storage *right now*.

Did you look at those benchmarks? It doesn't give much more performance over a regular laptop drive. Not even enough to best a desktop drive, like a WD Black.

For small file read/writes, SSDs are still magnitudes faster. (20-50x faster than this drive?)

At best, you could say this drive has built-in readyboost.

Re:Or wait..

[InfoJunkie777]

Just looked up 1 TB SSD on this OZC you mentioned. The costs must be going down. Listed at ONLY $4,000. Still a bit pricey, but coming down. I remember reading the first release of the 1 TB SSD drive at $10k.

Re:Or wait..

[Anonymous Coward]

You are correct, though you only need one form or the other most of the time.

When you're talking about measuring or counting out a substance using some known measure or scale, then you have to use an actual numeric figure such as 25 or 100, or a word or phrase that can stand in place of that number, such as "few", "several", "many", "dozens", "hundreds", etc., as the situation warrants.

A block of data is a collective "substance" that can be measured by the GB (among others), so using the hard drive example:

"This disk holds many more GB of data than the old one."
"This disk holds 100 more GB of data than the old one."

If you're talking about totally vague amounts of some substance, without any kind of divisions or measure (or where a commonly accepted measure is being intentionally ignored), then you'd most likely use "much", or words like "somewhat", "lots", "tons", "way", or perhaps one of the more colorful slang expressions. This is especially true if the sentence is still grammatically correct when written without the extra word, as in the case of the above hard drive, but without talking about the exact size:

"This disk holds more data than the old one."
"This disk holds somewhat more data than the old one."
"This disk holds much more data than the old one."
"This disk holds boatloads more data than the old one."

(Roughly in order of increasing impact. Note how the last sentence uses a common slang term.)

TL;DR:

You can use either form with almost anything that exists as a collection that can be counted or measured in some manner, it just matters whether you mention the actual unit of measure (if you do, use a number or "many", "several", etc.).

Thus endeth our English lesson for this week. :-)

Re:Or wait..

[Eric S. Smith]

It would have been correct if he had said "many more datum", but that's just weird.

No, that would be like saying "many more apple" — "datum" is emphatically singular; it's "data" that people like to play fast and loose with.

The opposite case of "dice" might be of interest to ESL learners: the singular form is "die," but many people say "a dice." Are they ignorant of their mistake or are they speaking a newer dialect of English? Probably.

Re:Or wait..

[Anonymous Coward]

"10TB SSD in your laptop and never have to worry again "

I remember thinking that with my 10mb XT.

Re:Or wait..

[Firethorn]

This would work for most people and I predict 90% of desktops will be SSD-only in about 2 years.

Call me a pessimist, but personally I figure that this will take at least 5-10 years. My 2 year figure would be for 50% of laptops being SSD, and I haven't seen any news of SSDs breaking double digit laptop penetration yet.

My reasonings for laptops: 2.5" Drives are more expensive per gig than 3.5", conserving power is a good thing for laptops, and shock resistance is a selling point - basically extra concerns give ssds more advantages in a laptop form factor, whereas in a desktop a few extra watts isn't going to kill you, shock isn't normally a worry, etc...

You're not going to see a 90% penetration until a basic SSD that meets requirements is cheaper than the cheapest HD that meets requirements. This will be difficult as long as manufacturers can get a 80GB HD for like $20.

Remember that for most people price is king, performance is secondary. Then look at that flash is currently like 20 times more expensive than HDs, and is only doubling in capacity per dollar between every 12-18 months. That's 5 doublings left to beat 'current' hard drives.

Once you can get, say, a 100 GB flash drive for cheaper than a 200 GB HD, then I figure the 'base line' units will switch over to SSD, with 'power users' simply installing an ancillary spinning platter - perhaps a 10TB unit?

I know quite a few 'less technically' inclined people running media servers today, downloading tons of HD video, etc... But I also know people who still haven't busted 8 gig, and windows updates take up more space than their my documents + extra installed programs.

Re:Or wait..

[Anonymous Coward]

Not a native english speaker either, but I thought that was about "mass nouns", i.e. nouns that refer to some more or less undefined large number of things as a whole.

RTFA

[copponex]

The trend continued in the remainder of the PCMark Vantage tests. On the first run of the benchmark, the Seagate Momentus XT performed surprisingly well. By the third run though, performance had increased dramatically and approached the level of the true SSD.

We should also point out that we ran this test numerous times, and after the third run, the additional performance increases stopped, which is to say Run 4 performed like Run 3. The screenshots of the actual Vantage performance summaries are available in the image gallery at the bottom of the page for those that would like to see the progression from Run 1 through Run 4.

So, it is slightly more expensive than a high performance disk drive, and offers most of the performance of an SSD. Most room on hard drives is taken up by massive media files, which do not need to be accessed at top speed because they are usually streamed for playback.

Eventually the the best drives will allow you to designate a folder for SSD storage only. Video editors should be able to buy a 1TB/32GB SSD drive and have a folder for the files they are currently editing. This may not be necessary if the drive intelligently identifies open files and transfers them to the SSD portion.

And I don't think this is just a big cache. I'm pretty sure hard drive caches disappear upon reboot.

Re:RTFA

[Kjella]

I did read TFA, and I have no clue how they reached that conclusion. Yes, it significantly improved on regular laptop HDDs but it didn't come close to touching the real SSD. In the IOPS test it was always behind by a 40:1 margin. If you look at the throughput benchmarks the Momentum XT usually outdid the regular HDDs with 3-4x, but the SSD outdid the momentum again with 3x. To take one example with the application loading: Fastest non-flash drive: 5 MB/s, Momentum XT: 20 MB/s, Patriot SSD: 60 MB/s.

That and in this case they clearly maxed the performance running the one application (well, test) over and over again. What happens if I go play a game, open my mail, write a document and open my browser again? Will it again be dog slow filled up by all this other junk? What happens if I delete a much used application, will the movie I download to that spot keep taking up cache? I mean if they start with a fully pristine disk where all usage counters are at zero and they started running this one application, of course it'll all be cached very quickly because nothing else is in use. But that is like the ultimate ideal case, and not even that one is very impressive...

Re:RTFA

[copponex]

That and in this case they clearly maxed the performance running the one application (well, test) over and over again. What happens if I go play a game, open my mail, write a document and open my browser again? Will it again be dog slow filled up by all this other junk?

I don't know... what happens if you whinge for eternity about products you don't plan on buying? If you think that practical application benchmarks like FutureMark are lies, then you have two choices: buy the product and conduct your own tests, or don't. If all you needed were synthetic measurements and manufacturer claims to make your buying decisions, then feel free to spend $1500 on a 400GB SSD instead of $150 on a 500GB hybrid.

Re:RTFA

[Firethorn]

And I don't think this is just a big cache. I'm pretty sure hard drive caches disappear upon reboot.

Current caches disappear on loss of power because they use volatile ram.

If the file is still present on the spinning platter it's still a cache, just a non-volatile one, which is a good thing when it comes to speeding up boot times and such.

Re:RTFA

[Rockoon]

If you think that practical application benchmarks like FutureMark are lies

Actually, he is pointing out that he *KNOWS* that in this case, the TFA "conclusion" is lying. Thats a bit different than "thinking" that its a lie.

Now, what happens when a real shitbag fuckhead like you replies all the time to people they didnt even bother to read, in the defense of things that are obviously wrong?

Re:RTFA

[copponex]

The Seagate Momentus XT produced some interesting performance results. In the purely synthetic tests designed to look at raw throughput, the Momentus XT generally performed like a high-end, 7200 RPM hard drive--transfer rates typically fell between the WD Scorpio Blue and Seagate Barracuda desktop drive. In the system level benchmarks like those employed in PCMark Vantage, however, once the Momentus XT's adaptive memory technology learned the usage patterns and copied over the most commonly accessed bits of data to the solid state portion of the drive, performance improved dramatically. The Momentus XT was never able to truly match the performance of a true SSD, but its performance when the adaptive memory technology was used was far superior to a standard HD...

...with the alpha drive we had enough time to experiment for a few weeks with the drive acting as a primary boot volume, with an full OS install, etc. It is difficult to say that all of the performance benefits we experienced were due to the drive itself, because Windows 7 also adapts somewhat to usage patterns, but the Seagate Momentus XT seemed to speed up the system, bit by bit, over time. Boot times, application launching, and web browsing in particular performed better after using the Momentus XT for a while, than when the OS was first installed

He did not read the conclusion, and apparently you didn't either.

I was going to think of a colorful way to describe your ignorance, but decided my time was too valuable to be wasted on such a simple exercise.

Re:RTFA

[Anonymous Coward]

All of this information points to an improvement in HD speed, however you tweak it. But does anyone on here see where you could save over 100,000,000 hours of wait time a day simply by keeping the OS loaded via SSD so that even starting up with the machine completely off, not in standby, you could basically have instant-on, or something really close to it? Just by doing it that way we could save electricity in the millions of dollars every month in the U.S. alone.
I can already hear someone say,"well, if you simply leave it in standby mode you can have about the same effect and electricity usage drops down to almost nothing". Fine, but on this forum most users know that tens of millions of computers have to be restarted at least once a day for various reasons. And that too would take what? A few seconds if the HD kept the OS in SSD?

SLC flash

[Animal Farm Pig]

According to the article, it's SLC flash. It should have many more write-erase cycles than MLC.

Prices here:

[olddotter]

I read through the article to see the prices:
"The initial three drives in the Seagate Momentus XT line-up will retail for $156 (500GB), $132 (320GB), and 250GB ($113). Those prices equate to roughly $0.31 to $0.45 per gigabyte, which puts these drive within striking distance of a standard HD in terms of price and much less expensive than any SSD."

This looks interesting enough for me to look into the compatibility with my MacBook Pro. Anyone with a MacBook Pro had sizing problems when replacing the hard drive? I imagine the size tolerance to be tiny in the modern MacBooks.

Re:Prices here:

[NatasRevol]

No, they have a flexible rubber insert to hold the screw heads & the drive. So there's definitely room for variation.

However, I can't imagine that they would be built outside of typical spec for laptop hard drives.

Re:Prices here:

[aztracker1]

laptop sized (2.5") drives are pretty much all the same, in terms of size and mounting points, both of my Intel SSDs actually have a spacer on the top so that it "fits" the right size. I put one in my new desktop in March, it was a huge difference. When I ordered my Macbook Pro earlier this month, I also ordered a 160GB SSD for it. My desktop has an 80GB one for it's main OS drive. I am amazed at the difference it makes, Win7 boots in about 15 seconds to the desktop (from bios hand-off), and the Macbook makes it in about 8-9 seconds.

Re:Prices here:

[olddotter]

Thanks for your details. I have wanted an SSD, but I can't justify the price for a "hobby" machine. This seems like it will make a great compromise, and 4GB cache will probably go further with OS X than Windows 7.

A glaring omission by Seagate

[Lucas123]

When I read specifications on a drive, one of the key things I look for is IOPS and read/write speeds. But Seagate seems to have omitted that. I wonder why?

Re:A glaring omission by Seagate

[DaveV1.0]

They haven't been omitted. You need to read the second link. IOPs and read/write speeds are about 1/3 and 2/3 of the way down the page, respectively.

3rd run

[Anonymous Coward]

Throughout the article, the reviewer praises the abilities of the hybrid drive--after the 3rd run, which provides the drive with enough data accesses for it to predict what data will be checked next and pre-load it into the solid-state memory. However, on the first run, and on larger operations, the drive performs just like the normal 7200 rpm drive that it effectively is.

This hybrid drive certainly is a jump up over traditional HDDs, but I'm glad they benchmarked the drive against a true SSD as well--a quick glance over the graphs and the pure SSD seemed to be about twice as fast as the hybrid drive on several occasions. I chuckled when I first saw the graph on this page:

http://hothardware.com/Articles/Seagate-Momentus-XT-Solid-State-Hybrid-Preview/?page=7

Point is, if you're splashing out money to get a faster drive, get a "budget" SSD, not a budget-conscious traditional hard drive. Even the budget ones are miles above a standard or even a hybrid HDD, and the first-generation issues have been resolved (the disk controller problem being the main one) as far as I am aware.

For the same money you're spending for this hybrid drive, you could just get a massive standard high-speed drive and not be crippled for space. Or if you are spending the money, get a smaller-sized SSD and put your data files (*that's* why you have a 1TB drive currently, your 200GB mp3 collection and 600GB of "other" files) on an external drive, where speed isn't as much of a concern as web browsing or FPSs.

I went for the latter option: I picked up an Intel X25-M, and it's unquestionably rejuvenated a *4* year old laptop for daily use. It's faster than my friends' computers that are half that age, and I fully expect to get a couple extra years of service out of it because of that upgrade.

And some of the other graphs aren't very fair to SSDs, either--they're done on a logarithmic scale instead of a normal one. Transferring 200 MB/s is twice as fast as 100 MB/s, not something like 20% faster (which is what it seems at first glance). The continued dominance of the SSD in the tests is minimized by the choice in graphing.

Re:3rd run

[Bengie]

"Point is, if you're splashing out money to get a faster drive, get a "budget" SSD"

Show me a "budget" 500GB SSD. They're not competing with speed, they're mixing speed and *size* and it has a pretty good lead on other 500GB traditional HDs.

I could see a 2TB HD with 16GB of flash being really nice to RAID for storage. ZFS + RAID-6z + 8 of these + small SSD added to the RAID pool for caching.

But still, I can't wait for cheap 500GB SSDs with great (200MB/sec+ and 10k IO/sec+) performance.

Re:3rd run

[petermgreen]

Some of us for whatever reason want a lot of storage on our laptops. Generally on laptops the fastest HDDs you can fit are 7200 RPM and you can only fit one drive. Yeah messing arround with external drives is a possibility but it's one I'd rather avoid for a number of reasons.

That is what I see as the niche for these drives. Those who want a faster drive in their laptop without sacrificing storage capacity and without spending $1400 to get a SSD that is approaching the largest laptop HDDs in size.

Software / OS hacks

[rwa2]

Seems like you could do better if you simply could reorder the files on your traditional hard disk so that you'd get 100% readahead buffer hits. If properly optimized this way, your traditional hard disk should always be transferring near the max block read rate of ~100MB/s

I'm guessing this is what some of the boot profilers / optimizers are doing.

The readahead utility used by Redhat / Fedora (and also available for Debian) gives you some benefit loading lots of small files from the disk by reordering reads by inode number to minimize head seeks. The next major benefit would be if it could actually reorganize all those files into a single tarfile, and maybe even compress it a bit, so it can do a single large block read to get all that content off of disk and into RAM cache.

Re:Software / OS hacks

[rwa2]

Hmm, looks like one of the ClamAV developers was proposing something like this in 2006:

http://us.generation-nt.com/answer/patch-rfc-rearranging-files-improve-disk-performanc-help-181277711.html#r

Could probably use blktrace to do the profiling by logging inode access patterns and identifing inodes that are frequently accessed together. Then those inodes could be packed / defragged next to each other on disk.

BTW, here's an interesting utility to plot inode accesses recorded by blktrace/blkparse :
http://manpages.ubuntu.com/manpages/jaunty/man1/bno_plot.1.html

Also this looks similar:
http://oss.oracle.com/~mason/seekwatcher/

This is the wrong place for this optimization

[Omnifarious]

Filesystems have a much better idea of what data is going to be used frequently. This is an optimization they should be making. Seagate can make some good guesses by looking at block-level IO statistics, but that's like trying to optimize bytecode, all the really useful information is gone by the time you get to block-level IO.

I think hardware vendors should be supporting more interesting experimentation on the filesystem front instead of coming up with proprietary hacks like this that are basically a half solution.

Re:This is the wrong place for this optimization

[inKubus]

What if this drive could show up as two devices and a control driver and the driver allows a very fast copy between the two (without using SATA to the mobo)? SCSI has control drivers, used for scanners, tape libraries, etc. It wouldn't be to hard to graft a few controls into this drive and then have out of band front-line to nearline migration that happens on the disk autonomously.

Re:This is the wrong place for this optimization

[hatemonger]

From the benchmarking, you're right. It's a faster HDD than other HDD's, but nowhere near the performance of an SSD. Labeling it as an SSD/HDD hybrid is misleading and wrong.

Re:This is the wrong place for this optimization

[Omnifarious]

Oooh, that would be really interesting, and a very nice feature to have when you're writing a filesystem. :-) I was thinking something similar, but didn't have the idea of the control driver, and that's a very nice addition.

Re:This is the wrong place for this optimization

[m.dillon]

I would go as far as to say that Seagate is chasing an impossible dream here. I've done extensive tests with HAMMER on DragonFly with a SSD caching HD content.

First of all, 4G of flash will not do diddly to improve the performance of a high capacity hard drive in the real world. The minimum you would need is 20G to have any chance of being able to cache filesystem meta-data. The 40G or 80G Intels fit the bill very nicely.

Secondly, it absolutely matters what data the system decides to throw onto the SSD, and there is no way a hard drive can figure that out from short-term activity logs. The HD would have to analyze several days worth of fine-grained activity to get a clue and it is clearly not going to be doing that. Operating systems and filesystems have a much better clue in this matter.

-Matt

Re:This is the wrong place for this optimization

[aztracker1]

Agreed, I'd recommend 64-80GB SSD for main OS drive, and a larger one for media storage, now if I could get even 40GB SSD and 320+GB of HDD showing up as even separate drives on the same device that would be nice. I'm using about 45GB of my 80GB intel SSD on my desktop, and my laptop is using a bit more, (because my VMs re on the same drive).

Re:This is the wrong place for this optimization

[Bigjeff5]

Except you can't put out something new and be able to sell it to 99% of the market.

I'm sorry, that's not how business works. You don't just play with something without some plan to get it to market.

The way to do this is as follows:

Phase one:
Create a hybrid drive that is significantly faster than traditional HDDs (though still slower than SSDs) without any of the storage problems of SSDs. Vendors copy you, pretty much all HDDs on the market are eventually hybrid.

Phase two:
While this is booming, be talking to Microsoft and Apple about how much better it would be if the OS were controlling the cache instead of the hard drive. You could suddenly have the fastest OS on the market with any drive. Dollar signs flash, and the next versions of the OS (or some update, if that's feasable) is capable of doing the caching.

Badda bing, badda boom, success. Microsoft and Apple aren't going to invest in something like this unless there is a market for these drives. There currently isn't, so Seagate needs to create the market first. THEN it can sell OS-handled cache.

Re:This is the wrong place for this optimization

[Pascal Sartoretti]

Filesystems have a much better idea of what data is going to be used frequently. This is an optimization they should be making. Seagate can make some good guesses by looking at block-level IO statistics, but that's like trying to optimize bytecode, all the really useful information is gone by the time you get to block-level IO.

You are 100% correct, but this drive has the huge advantage of working with ANY file system and ANY operating system.

Re:This is the wrong place for this optimization

[Omnifarious]

Or, how about this instead?

Phase one, release SSD drives that are clearly faster and make a bunch of money from early adopters who think they can use them.

Phase two, listen to developers who are trying to make them work better. Implement things like the 'release' command. Offer an idea or two of your own, like the controller side copying.

Phase three, release the new version of the drive that supports all of that stuff and make even more money.

Your version relies on back-room deals with proprietary software makers and will probably ultimately result in a worse solution for everybody.

Both versions make a whole bunch of money for the hardware manufacturer. Your version treats users as passive idiots who haven't a clue. My version treats them as active participants in the process hardly worthy of the word 'user'.

Different partitions, not as good?

[Firethorn]

Performance is better than traditional drives in almost all benchmarks, and reaches up to 2x the performance when using commonly accessed files (like, the operating system).

I agree, and I think the 'different partitions' of the parent poster isn't as good of a system as an automatic cache system.

Consider that even for linux installs, not all system files are going to be equally read or written, same with applications and even data files. Some will be used almost all the time, some will collect virtual dust for years.

Newegg prices:
500 GB laptop drive: $65. 13 cents per gig
65 GB SSD: $145. $2.23 per gig. (Went 64GB to avoid small cap premium).

At around 6% of the cost, it's essentially 'free' to duplicate the SSD data on the actual platter. You're paying 52 cents for that 4GB on the platter, $8.92 for the SSD. That puts flash very much in the price range for cache - where you're willing to spend the money for the extra performance, but just can't afford to put everything on there.

I'm sure an 8GB version would make the geeks very happy, but I wouldn't be surprised if 90% of my most commonly accessed files would easily fit on 4GB. More is better, of course, but at 20x the price, how much 'more' is really better?

With smart caching, I wouldn't be surprised if doubling the storage had less than 10% difference in real world performance.

The main reason I'd want 8 or even 16GB is that I recently went to a SSD for my main computer and I LOVE the fast hybernate(and wake up). So there's 4 or eventually 8 GB right there for the hybernate file. Don't forget the whole GB vs GiB debate... Memory uses 'real' gigabytes, which means that you'll likely need 5GB as HD/SSD makers measure them...

System drive / storage drive?

[l0xin]

It sounds pointless to me, however the manufacturers obviously see padding out the transitional market as a potential reward generator. But really, I got an 80GB Intel X-25M (one of the highest performing SSD's) for £120 at the beginning of the year. That's really not expensive for the performance increase / noise reduction / energy saving boost it gives (which is fairly awesome), and 80GB is easily enough for me to use as a system drive, leaving an HDD as storage. It wouldn't surprised me to find out that a lot of users that shun an SSD for a larger capacity HDD as their system drive, never actually USE the extra space.

Huh?

[Whispers_in_the_dark]

FTA: "To put it simply, the most commonly accessed data on the platters get's copied to the much higher performing, SLC Flash memory, which results in a performance boost." Read more: http://hothardware.com/Articles/Seagate-Momentus-XT-Solid-State-Hybrid-Preview/?page=2#ixzz0orEbgttB This makes no sense to me -- that would seem to imply the most likely thing to end up on the SSD is my swap partition, which is the last thing I want on SSD. Yeah, read would be faster, but the wear would be awful. Maybe I'm missing something. I'd probably be happier if it just exposed the 4GB as a different partition...

Re:Huh?

[EmagGeek]

Actually, the whole point behind ReadyBoost was to put the swap data on a flash drive because it gave such a large perceived increase in system performance. Doing thousands of tiny random read/writes to even a slow USB flash drive is better than doing it to a platter...

Re:Huh?

[gnasher719]

his makes no sense to me -- that would seem to imply the most likely thing to end up on the SSD is my swap partition, which is the last thing I want on SSD. Yeah, read would be faster, but the wear would be awful. Maybe I'm missing something. I'd probably be happier if it just exposed the 4GB as a different partition...

Let's do a bit of math. Let's say we are writing 100 MB/second, 24/7, 365 days a year. A year has about 30 million seconds, so we write about 3 million GB. Since the drive has 4 GB, every byte is written 750,000 times. That should be no problem at all, I'd expect the drive to last a few years at that rate. And nobody writes 100 MB/second 24/7.

But could we just say that if you spend significant amounts of time writing to your swap file, then you don't need a hybrid drive, but more RAM?

No, not really

[SmallFurryCreature]

The way to get both performance and storage right now is to by TWO disks. An amazing concept I know. Who would have thought it was possible to get more then one HD/SSD into a PC.

Every single story about SSD's seems to bring out the idiots who want everything on one disk. Good thing these guys ain't farmers or they would be trying to plow the field with a Ferrari or cruise town with a tractor.

This drive is only of use to people who can't afford a real SSD and are limited to a laptop with only one drive bay and even then you would get far better performance with a normal SSD and an external drive for your porn collection.

Yes yes, there are people who use a laptop AND have need for far bigger datasets but on the whole, those people also need far greater access speeds then a traditional laptop HD can offer. I find it amazing to see someone claim he needs to edit video on a laptop with a 500gb 2.5 inch HD running at 5400 rpm. Who are you trying to kid?

And this drive won't be much help here. 4GB is just a cache file, if you are lucky it caches the right files but if you are doing complex stuff these "smart" caches often get horribly confused and start caching the wrong data. Like Vista trying to cache torrented files. Yes, I know it accesses the file a lot but please don't try and cache a 10gb file on the same HD. What's the fucking point? If you for instance will be running a large database from this drive, I am willing to bet its cache performance will degrade as it simply has to much to cache. Small caches only work when a small amount of files is requested a lot and the rest isn't. Like a porn collecton on your OS drive. Video editing, databases, filesharing always screw up caches.

If you really want performance in a laptop, spring for one with two drive bays, put as much memory in it as it can hold and get an SSD and a HD. A real SSD not one of the cheap ones some laptop companies put inside. An SSD is NOT just a fast HD, they truly are in a class of their own. And even if you got only a small single SSD, then you can still save space by putting your music/porn on a flash card or usb stick instead.

I wonder if people can ever get it into their heads that an SSD is about speed, not about capacity. Then again, since every single netbook these days comes with a 360gb slow ass HD instead of small but fast SSD, I think I might be fighting a loosing battle. Seems the average customer can only judge something if the number is bigger.

Re:No, not really

[dingen]

The way to get both performance and storage right now is to by TWO disks. An amazing concept I know. Who would have thought it was possible to get more then one HD/SSD into a PC.

In most computers sold today, the fitting of more than one harddrive is not possible. Besides that, it's a very difficult to manage solution, as people will have to manually decide what to put on the fast drive and what to put on the large drive. All in all it's a very fiddly solution, only available to tech-savvy folks with customazible computers. Not to mention the fact that two drives are more expensive than one.

In the real world, a hybrid drive such as Seagate is proposing is a lot better in almost every way thinkable. It's just one drive, so it will fit in basically every computer in existence and it functions completely automatic, as the user is presented with just one storage medium. The tests in the article prove this type of drive is both faster than traditional drives and a lot cheaper than SSD's, so it really is best of both worlds.

I wonder if people can ever get it into their heads that an SSD is about speed, not about capacity.

That's because harddrives are meant for capacity, not speed. Nobody thinks "Hey, my computer is slow, lets get a new harddrive". People buy harddrives to store their stuff on, so they want the drive that will hold the most stuff. So if you want to sell a lot of harddrives, you have to make sure they are able to hold a lot of data first and then think of a USP on top of that, which is exactly what Seagate is doing by creating these hybrid disks. The result will be large and fast disks for everyone.

Re:No, not really

[Bigjeff5]

4GB is just a cache file, if you are lucky it caches the right files but if you are doing complex stuff these "smart" caches often get horribly confused and start caching the wrong data.

You do realize that the reason your computer is so fast is because of progressive layers of cache, right?

The fastest cache on the system is L1 cache. It's also the most expensive. Next is L2 cache, which runs at about 1/10th the speed of L1, but it's much cheaper and so there can be more of it. That it's only an order of magnitude different means the larger L2 cache has time to fill the L1 cache before the L1 cache is completely empty. Then comes L3 cache (usually), which is again about 1/10th the speed of L2, and it keeps the L2 full. Then RAM, which has kept pace pretty well and is about 1/10th the speed of L3 and keeps L3 full. And here is where things break. RAM speeds are measured in nanoseconds. Spinning disk hard drive speeds are still measured in milliseconds, and not even 1 or 2 milliseconds, more like 5-10 milliseconds. That's a couple orders of magnitude slower and breaks the chain of cache that we had going, and it is not enough to maintain full RAM at all times. What we need is a cache that is about 1/10th the speed of RAM to sit between RAM and Hard Disk.

SLC Nand flash, with its sub-millisecond read and write times, fits the bill perfectly. It's basically a scaled up version of the caching they use on hard drives already, and because of its size should be much, much more effective.

Re:No, not really

[imgod2u]

In the real world, a hybrid drive such as Seagate is proposing is a lot better in almost every way thinkable. It's just one drive, so it will fit in basically every computer in existence and it functions completely automatic, as the user is presented with just one storage medium. The tests in the article prove this type of drive is both faster than traditional drives and a lot cheaper than SSD's, so it really is best of both worlds.

It's not actually the best of both worlds. It's more along the lines of something good from both worlds. First, this only improves read speed as it's not used as a write-cache. Second, random read and write speeds are just as abysmal as traditional HD's. And lastly, it's not quite going to match the price/capacity of traditional drives due to the need for mult-GB of SLC cache.

Now, it's certainly better than a traditional drive at sustained reads and offers better price/capacity vs SSD drives. But I get the feeling that a pure hardware, blind algorithm isn't the best way to go about this.

I would've preferred it if the drive presented the flash cache as a ReadyBoost drive to Windows and have the OS manage what needs to be cached. Certainly Windows knows more about which of its own disgusting innards needs to be readily accessible better than a hardware algorithm.

Re:No, not really

[dingen]

I would've preferred it if the drive presented the flash cache as a ReadyBoost drive to Windows and have the OS manage what needs to be cached. Certainly Windows knows more about which of its own disgusting innards needs to be readily accessible better than a hardware algorithm.

Microsoft would like this too, as they presented this idea in 2006 already with the "ReadyDrive" concept. But it hasn't caught on, mainly because of a lack of Vista adoption.

Re:No, not really

[MobyDisk]

First, this only improves read speed as it's not used as a write-cache.

Are you sure that is how they implemented it? That is very foolish of them. That means this won't help my compile times.

And lastly, it's not quite going to match the price/capacity of traditional drives due to the need for mult-GB of SLC cache.

But it is darned close. I saw that this drive is only a $40 premium over existing drives. That sounds very worthwhile on a mid to high-end computer.

Re:No, not really

[Anonymous Coward]

The world is going more and more mobile. Can your netbook fit both an SSD and an HD? No it can't. Can most notebooks? No they can't except when optical drive deleted.

Now, after a couple years of SSDs going down in price much less than people had antipicated, is the best time yet for manufacturers to be introducing hybrid drives.

Re:No, not really

[master811]

Second, random read and write speeds are just as abysmal as traditional HD's.

That's where you're wrong, according to this, the random reads at least on the smaller files are significantly faster than a normal HDD.

http://www.overclockersclub.com/reviews/seagate_momentus_xt_500gb/4.htm

Re:No, not really

[Anonymous Coward]

until one of those two disks fails.. and you're left with half of your data

Re:No, not really

[imgod2u]

That's where you're wrong, according to this, the random reads at least on the smaller files are significantly faster than a normal HDD.

http://www.overclockersclub.com/reviews/seagate_momentus_xt_500gb/4.htm

I was going by Anandtech's results:

http://www.anandtech.com/show/3734/seagates-momentus-xt-review-finally-a-good-hybrid-hdd/3

Of course, if you go in 1MB files, it's going to be faster (assuming the file isn't fragmented).

Re:No, not really

[imgod2u]

A more mature of that "ReadyBoost" exists in Windows 7, which has caught on; a lot. I imagine that something like it exists in Linux and OSX (or will if drives exist).

The OS will know enough not to thrash the flash cache with swapfile operations.

Re:No, not really

[BikeHelmet]

I wonder if people can ever get it into their heads that an SSD is about speed, not about capacity. Then again, since every single netbook these days comes with a 360gb slow ass HD instead of small but fast SSD, I think I might be fighting a loosing battle. Seems the average customer can only judge something if the number is bigger.

We should list 4KiB read/write speeds on the box!

I just tossed together a DIY laptop for someone that required a desktop-replacement laptop that is easy to move/carry. What did I choose?.... 500GB 7200RPM 32MB HDD. Why?... their old laptop had an 80GB drive completely full. 100+ GB SSDs are prohibitively expensive.

The sad thing is, all our data is getting bigger. Pictures used to be about 1MB. Now it's not uncommon for them to be 6+ MB. OS's used to be about 2GB. Now they're at least 12GB. We easily need 6x the storage that we used to.

Re:No, not really

[Anonymous Coward]

Good thing these guys ain't farmers or they would be trying to plow the field with a Ferrari or cruise town with a tractor.

Ferrari have made tractors for some time http://www.ferrari-tractors.com/
So have Lamborghini
http://www.lamborghini-tractors.com

Re:No, not really

[Firethorn]

It's more along the lines of something good from both worlds.

Sometimes this is better than getting the best of one.

For example, this sounds perfect for my 'games' drive. I tend to install a game, play it intensively for anywhere from a week to a couple months. After that, it tends to go on the backburner, but I might play it every other week or so.

Given the size of games today, I don't want to effectively be paying 50% more for the game to keep it installed on a SSD. It can be a pain in the butt to 'move' an installed program such as the game. I DO like the extra speed offered by SSD, but can put up with slower loads, especially for a game I haven't played in a while.

Heck, don't forget that it's not typical that all files installed are frequently accessed, whether it be game, os, or other program.

On a cost basis, I'd love to see a 3.5" version of this, 1-2 TB of storage, 8-16GB of SSD. Keep my OS and programs on the SSD, move my games to the new combination drive. Don't have to worry about doing file links to preserve my 64GB SSD.

Durability and Power

[Johnny Mnemonic]

This drive still suffers from the historical bugaboo of spinning platters: it is damaged by shock. Also, it has the power draw (and heat output) of other spinning media.

Those are the two biggest reasons for SSD, especially in notebooks. Performance improvements are a factor, but I think they're the least interesting. In this respect, Seagate still needs to bring an answer, and they need to do it fast to justify their run up in stock price.

Re:Durability and Power

[MobyDisk]

The platform is the benefit though. Right now it has 4GB/250GB is 1.6% flash. Once this proof-of-concept works, I bet they could make one with closer to 20% flash. At that point it might spin-up the platter drive rarely enough that the power-draw issue goes away. If the drive is usually parked, the shock benefits improve a bit. That might be a good-enough solution to stick around for 5-10 years before the next thing comes along.

Response time

[Anonymous Coward]

The response time seems to be just as crappy as a ordinary hdd. Horrible.

Why a log scale at HotHardware?

[fljmayer]

The logarithmic scale in the graph at the HotHardware link is really confusing. I think a linear scale would show the differences between the drives much better.

Re:Why a log scale at HotHardware?

[hydromike2]

seriously, the patriot drive out performs the seagate hybrid third run by a factor of 2-3 in everything except the 'media center' benchmark. What appears to be a 10% difference if you dont look at/or understand the scale between the patriot and seagate 7200.11 is a factor of 2 difference. In short the graph is bullshit trying to make the seagate hybrid look comparable to the patriot.

MOD PARENT UP!

[Anonymous Coward]

Wow!

Samsung already did this.

[lazn]

I have one of the old Samsung drives from 2007, a 80GB with 256MB of flash.
http://en.wikipedia.org/wiki/Hybrid_drive
http://www.samsung.com/us/consumer/office/hard-disk-drives/hybrid-hdd-flashon/HM08HHI/index.idx?pagetype=prd_detail

You lost me...

[Joce640k]

You expected a built-in feature of the Windows OS to run on Linux?

Re:You lost me...

[dingen]

No, I would expect a harddrive to work in Linux. A harddrive which relies on ReadyDrive would not be a very good product, as it would only work correctly in Windows. That's why those type of harddisks never caught on, even though Microsoft did try to push this concept.

What Seagate is doing now, is using the ReadyDrive-concept of hybrid harddrives, but provide ReadyBoost-type technology on the controller of the harddisk instead of relying on the operating system.

Re:You lost me...

[teg]

No, I would expect a harddrive to work in Linux. A harddrive which relies on ReadyDrive would not be a very good product, as it would only work correctly in Windows. That's why those type of harddisks never caught on, even though Microsoft did try to push this concept.

It would be a good product for the vast majority of laptops sold. Not being linux compatible is not why it didn't catch on as most people unfortunately don't care. There is a huge market for Windows-only hardware and software. The reason is probably as simple as the cost/benefit ratio not being what it should be, or the race to the bottom direction that has dominated the PC market with the exception of Apple.

Just go grab 4 of these...

[ub3r n3u7r4l1st]

and run then in RAID 10, that beat budget SSD at any time with a large storage for applications. No longer worry about "pre-configuration" in Windows. Just install and run.

Worst of both worlds?

[Hythlodaeus]

Not only would this still have the noise of a physically spinning platter, and still have long seek times for anything outside the cache, but it would also wear out the rewrite capacity of the flash part with frequent cache filling.

Re:Worst of both worlds?

[Anonymous Coward]

It's SLC flash = several million erase cycles per block.
the flash part is used as a large read cache, so data needs to be transferred from platter.
assuming 5ms for 4kB, that's 800kB/s or ~69GB/day.
even with 'only' 1M erase cycles, we're talking about 4PB, or about 150 years of 24/7 reading...

Re:Worst of both worlds?

[VinylPusher]

As an SLC part, the wearing out bit of your argument falls down.

However, what should be a slight concern is the issue of having two points of failure in one device. If either the flash or the spinning disk dies, the whole drive is screwed. It will create nightmares for anyone sending a failed drive off for data recovery.

If Seagate have posted an MTBF figure for the drive, I'll be surprised if it's accurate.

piness

[Anonymous Coward]

I think AC is talking about Dr. Robotnik's hapPINESS.

yeah well 4gig is all used up frequently

[Anonymous Coward]

with a couple of different compiler environments. matlab and mathematica, for linux. chromium, etc. my 4gig is all used up frequently, and after a couple of days of leaving the laptop on and just suspending, i find a couple of gig of swap gets used too.

Hindering progress of SSD

[Twinbee]

It's a nice idea, but it'll hinder the progress of SSD if it catches on. I'd rather SSD replace HDs completely.

Logarithmic Chart Axis: Confusing!

[grondak]

Logarithmic chart axis scales are used to make logarithmic functions easier to draw by hand-- when you show a logarithmic function against a logarithmic axis scale, they present as a straight line instead of a curved line. Zip, the straight edge does all the work to make the result very nice, thank you Mr. Straight Edge and Mr. Light Table. Logarithmic axis scales are also used in the case of a computer-drawn chart of a function formerly/typically shown with logarithmic axis scales when drawn by hand. This disc performance chart is neither! It's simply confusing, because the slower drives' functions appear to be bigger than they normally would in comparison to a linear scale. Does any other site or graph show disc performance data on a logarithmic chart?

Re:Logarithmic Chart Axis: Confusing!

[treeves]

The logarithmic scale also allows you to see the difference between 3 and 5 when the data also includes numbers as high as 167.
A linear scale can't do that: if it includes 167, 3 and 5 will be indistinguishable.

review is fail

[strstr]

The numbers don't tell much. Part of adding flash cache is supposed to be so frequently used data can be accessed with much lower latency and more randomly, and it's not aiming to speed up a single application but the entire system. It would have been nice to know how the desktop performance improved or "felt" - overall responsiveness of OS and application loading, multitasking, etc. That's where the performance improvement is going to be.

glorp

[seifried]

undoing an accidental moderation

Other solutions? How fast is it really?

[scottwilkins]

I wonder if just putting 4 Gig of super fast cache memory would have the same affect, or at least something similar? I'm going to wait for the high end drives like this. The one's that will have the faster flash memory. Just like current SSD drives, some are faster than others. Sometimes quite a bit faster. Seems to me if the flash memory doesn't have much better throughput than the normal hard drive's sustained throughput rate, then it's not worth much except for seek time improvements. So I hope to see soon what the real throughput rate is for the flash memory part of this drive is.

What happens when you defrag the drive?

[MobyDisk]

I bet it is awful after you defrag it. The drive will see a lot of random reads and writes, which probably thrashes the Flash and randomizes the contents of the cache.

Lose vs. loose

[Anonymous Coward]

http://www.wsu.edu/~brians/errors/lose.html

Loose vs. lose

[Anonymous Coward]

http://www.wsu.edu/~brians/errors/lose.html

Second time in one thread... two chances to educate!

4gigs of ram - still too expensive.

[Firethorn]

4 GiB of RAM would still have to be filled up at boot, and flash is rounning around $2 a gig vs around $20 for RAM.

$.17 a gig for a 7200rpm laptop hard drive.
$2.23 a gig for a SSD
$19.50 a gig for the cheapest 2gig ram stick off of newegg.

Sure, RAM is a lot faster even now, but we're back to looking at various levels of cache - 10x the price for.

When you go looking at RAM, I'm going to say it's better off in the system, especially now that so many people are going to 64bit systems, and I've heard that Win7, possibly Vista will use excess RAM for caching. The hybrid drive proposed here is an interesting and I think well positioned device - nonvolatile cache means that the system doesn't have to worry about flushing a relatively huge 4gig of RAM to the disc in case of failure.

So I'm back to that I'd like to see a 'premium' product offering 8 or even 16 GB of flash for the 'caching' operation. That'd allow the commonly used parts of my OS, my most commonly used 2-3 applications, as well as the commonly used files for whatever game I'm playing.

Given the issues with installing STEAM games to different directories, for example, this as a secondary drive for my computer would allow me to install STEAM on it, enjoy my game of the month at (near)SSD speeds, the steam app itself would end up on the flash portion, but I wouldn't have to worry about flushing the game of last month off of the SSD, and I can still play it without downloading it or restoring it from backup if I want to.

16GB SSD + 500GB HD

[Anonymous Coward]

I'd like to see 2 logical drives in my laptop, a 16 GB SSD for my OS (Mac OS ~8GB) plus MS Office, Firefox, and a few other apps I use...and also a 500GB HD for my music, photos, movies, and large docs. Why is this so hard? Should be doable for a nominal increase in cost. I know it's possible to get laptops with 2 drive bays, but I want something small, something that probably only has one drive bay.