Seagate Releases Hybrid Hard Drive

An anonymous reader writes to tell us Seagate has released a new hybrid hard drive. This new drive adds the speed of a solid state drive to the conventional hard drive. Originally designed for laptops this new drive comes in 80, 120, and 160 GB flavors and features 256MB of flash memory.

Comments

Hmm

[PalmKiller]

Didn't samsung or some such outfit already do this?

Re:Hmm

[PalmKiller]

I found it, it was samsung over a year ago... http://www.infoworld.com/article/06/05/17/78429_HNhybriddrive_1.html?DISK-BASED%20BACKUP%20APPLIANCES

Re:Hmm

[cindysthongs]

Score -1 on Seagate then for being late to the party

Re:Hmm

[dbIII]

If I don't really care about the price I can order a 300GB 15,000rpm disk with 4GB of cache from IBM now. How does a hybrid disk improve on this?

Re:Hmm

[mindsuck]

It doesn't need to be spinning constantly since data can be cached on the internal memory so it improves battery life on portable devices such as notebooks where a 15k rpm disk would drain your battery pretty quickly.

Re:Hmm

[dbIII]

Good point - laptops and Vista stupidfetch and not just a cache. It still seems a little small to be able to allow the drive to spin down when the machine is in use but I suppose if it is on and idle the background processes might not be enough to fill the flash and wake up the disk.

Re:Hmm

[Anonymous Coward]

That IBM's 4 GB cannot be used to speed up booting, because it'll empty at shutdown. This hybrid can have (at least portions of) the bootup close to SATA wire speed (1.5 GB/s) from the 256 MB instead of it's disc read speed (44 MB/s in TFA) -- but it'll have to be very smart about what to keep in that 256 MB, of course.

But of course, that 15K monster is no laptop drive. (Although I wish there were 5K/15K selectable laptop drives -- I do use my laptop docked for longish periods, and 15K rpm discs have to be under 3" anyway, and fast random seeks matter to me and my laptopped database development quite a lot.)

A version of this hybrid with 1 GB or 2 GB would be very nice for super-quickly un-hibernating the entire session just where you left (for laptop use), without any hassle with separate Flash devices. And I do use the fingerprint scanner, thank you. :-)

Re:Hmm

[F4_W_weasel]

it will actually help you social skills, when you say that you have a HYBRID hard drive, it will make people think that you are environmentally friendly, the chicks will stop and listen to you.
Even if you say that the hard drive is just for porn.

Re:Hmm

[Thwomp]

Well if a hybrid drive doesn't spin as much isn't that pretty much true? I would assume power consumption would be less than a conventional hdd.

Re:Hmm

[sgartner]

AFAIK the flash is not used for "normal" last access caching, it must be used intentionally to cache part of the drives data. It appears from what I read that what is in the flash may not actually be stored on the physical platters at all (so the data is in one place or the other, though they appear to be one drive to applications). The upshot is that putting one of these drives into a computer that does not specifically have support for hybrids may not use the flash portion at all.

Someone please correct me if I've misinterpreted this.

These drives are great... but,

[GonzoTech]

I've seen one of these at a trade event in Atlanta earlier this year. The idea is great, and after much strenuous testing, seemed to still work great. I can't wait to get my hands one some!

Re:These drives are great... but,

[tomhudson]

If they make it at the same plant my 4 DOA drives came from, all I can say is "etter you than me."

Re:These drives are great... but,

[Dirk the Daring]

If they make it at the same plant my 4 DOA drives came from [slashdot.org], all I can say is "etter you than me."

You might want to check out your new drive out-- It seems to have some data loss.

This is Great

[MrCrassic]

However, why did they only include 256MB of flash storage instead of a larger quantity like 2 GB or so?

Many people who exercise smaller flash storage options get flash drives larger than 512MB, so was it really that much more expensive to bump up the available flash storage a little bit?

Regardless, I look forward to the performance benefits devices like these will provide.

Re:This is Great

[nrich239]

Agreed. I wonder if, as the space of the drive goes up, so does the flash space?
IE, 256 for the 80gb, 512 for the 120gb, 1024 for the 160, etc

Re:This is Great

[Corpuscavernosa]

If I'm not mistaken, the flash memory serves to save things in a more quickly accessible memory when your computer goes to sleep or hibernates, allowing for an extremely quick "awakening". Hence these were designed with the laptop user in mind.

Re:This is Great

[hey!]

I'm assuming (TFA is slashdotted) because the flash is used as kind of persistent cache. If so you could confidently defer a write onto the actual platter until you are doing other things in the neighborhood, confident that the data will be there if something goes wrong like a power failure or crash. I don't think it would do much for read caching, for which volatile RAM is fine.

Statistically, 256MB of pending sectors is probably enough to get most of the potential benefits from reorganizing writes to the platter. And if you sell a gazillion of these, a buck saved on each unit is a gazillion dollars of profit.

Re:This is Great

[Anonymous Coward]

I'm thoroughly confused now about using flash as a write cache. Isn't the write speed for flash something like 40MB/s for the expensive stuff, more like 10MB/s for the normal stuff? And don't SATA drives typically have better than 100MB/s write speeds? Are you typically writing the same stuff to disk over and over such that caching it makes sense? There is no predictive pre-fetch on a write either, as far as I know.

I see your point about not really needing flash for a read cache, since you don't need to keep it between reboots. But reading flash is still faster than from a disk, and since you don't need to keep it energized to maintain its state, you could have a cooler, quieter drive (and longer battery life) if you use flash instead of volatile memory.

Re:This is Great

[moderatorrater]

Actually, this is a persistent read cache. As mentioned in another reply, flash memory is much slower at writing but very fast at reading and it doesn't need to be spinning to be read. This allows for much faster boot times in laptops/etc because it doesn't have to wait for the disk to spin up.

Re:This is Great

[EvanED]

And don't SATA drives typically have better than 100MB/s write speeds?

That seems fast to me (copying one drive to another maxed out at about 40 MB/s, but I had an older SATA drive so maybe newer ones are faster), but that's for sustained transfer. If you're doing stuff like metadata operations, which are all journaled, it looks more like "write a journal block saying you're updating this bit of metadata, seek to the inode or indirect block you're modifying and write it, seek back to the journal and write a commit message". IRL these are batched, so the overhead isn't horrible, but there is still plenty of seeking involved.

In fact, ideally the drives would ensure that streaming writes don't push out the metadata from the flash cache.

That said, letting the drive run slower and stuff for longer life and better power consumption is very attractive. I don't actually know what exactly their goal is.

Re:This is Great

[hedwards]

No, flash memory can be quite a bit faster than that. Most of the time the limiting factor is something other than the flash. USB2 can only do 480megabits/s, and that is bursts, using something faster than that would be a waste. Even having something that can do 480 tends to be a waste as most of the time the transfer is much slower.

As for SATA drives they don't normally do 100MiB/s unless the information is already in cache on the driven. The flash memory is basically there to be a larger cache which is persistent across boots, allowing for the bootloader, kernel and a few essentials to be guaranteed a faster access time. Any additional items that go in there depend upon what the specific manufacturer specific algorithm does.

The size of the flash is like the size of cache on a harddisk, bigger isn't necessarily better. You could give a HD 30mb of cache, but if it is using a poorly designed caching algorithm, the difference can be nonexistent if the important stuff isn't in it.

In this case, 256 ought to be enough for present day computing. Even Linux is a fraction of that size, sure you could include a few utilities that regularly run during start up along with the kernel, but when you start to get beyond a quarter gig, you are beginning to get into things that run less predictably.

Re:This is Great

[OrangeTide]

Why sell 2G on the first version when you can get all the early adopters to upgrade next year when you offer 1G to 4G versions?

Re:This is Great

[Anonymous Coward]

Even with 2GB, firefox would overrun the memory.

Re:This is Great

[walt-sjc]

OK, my laptop has 2G of ram. Most modern laptops have at least 1G. While 256M is better than nothing, it certainly isn't much in terms of todays OS's and apps.

So here is what I would suggest... Put a card slot on there. Let me put in as much as I need. MicroSD cards are nice and small, but may be too slow, even the SDHC variety. I'm sure they could come up with something that would work well however.

Re:This is Great

[aliquis]

Quite obvious:
They figured that very few would be intrested to buy the product if it costed whatever amount if would cost more than a regular drive if it has 2GB of flash memory.
That was the first thing which springed into my mind and I thought 2GB of regular flash memory would add to much on the price, but it seems like it doesn't use the kind of flash memory we are used to but a more reliable one which cost more, see beneath:

"These first-generation hybrid drives incorporate only 256MB of NAND flash, a pittance in comparison with the 2GB USB flash drive you can buy today for $20. But the memory in the older drive is typically multilevel cell flash (MLC) as opposed to the more reliable--and more expensive--single-level cell flash (SLC) that hybrid drives, as well as SSD models, use. Manufacturers say they decided on 256MB of flash in hybrid drives as an entry point that wouldn't throw the cost of the product out of whack when compared with the cost of a standard 2.5-inch hard drive."
http://www.pcworld.com/article/id,138102-page,2-c,harddrives/article.html#

Re:This is Great

[MikeFM]

A hdd with a massive RAM cache seems like it'd be better than using flash. Or just get more RAM and use a filesystem that caches frequently read (and seldom written?) files to RAM? 2GB of extra RAM should cover the vast majority of files the typical user will use during a computing session - only forcing the hdd to work when a file needs to be read into cache or a file is written. I use such a file system on my Linux systems - I dunno how Windows handls such things. :)

Re:This is Great

[Born2bwire]

Maybe there was no appreciable benefit for having a larger flash storage. In cache, there are limits to how large you want to make the cache outside of physical or monetary constraints. You come into problems about the actual amount of memory that needs to be accessed at a time and the amount of time it takes to fill a block. If most of your data that you need has lengths of the two bytes, then it becomes less worthwhile to have a wide cache where you read in a block much longer than what is needed (you do not always want data in the same order as it is contiguously laid out in memory). This only concerns the width of the cache and is not a direct comparison to what you are asking, but it's been too long for me to gleam a better example. Having a large cache (or other lower level memory) is not very useful if you do not need to reaccess a lot of the data in relatively close intervals of time (temporal locality). In addition, the data you need may be located in many different areas in memory (spacial locality) requiring you to want to transfer smaller block sizes. Suffice to say, there are more considerations than just the size of your cache or secondary memories. The way in how memory is accessed and used determines to a large extent the size and configuration of the memory.

Obligatory

[TopSpin]

Have 'they' solved the problem of the limited number of writes a flash device device can handle. If it's only going to last a few months and then wear out I won't consider it! Pity the poor fool that forgets to turn off atime updates.

Re:Obligatory

[casualsax3]

Ha! I use Windows, I don't even have a noatime parameter!

Re:Obligatory

[Cyberax]

Actually, you have: http://msdn2.microsoft.com/en-us/library/ms940846.aspx (it's the first thing I do on a new installation of Windows)

Re:Obligatory

[arth1]

Beware that turning off the access time in Windows will cause defragmenters to be unable to organise files depending on access patterns, which leads to much worse defragmentation problems than if you leave it on.

Re:Obligatory

[TheRaven64]

I'm imagine they'd use this as a write-through cache. When you write data to the disk, it stores it in flash. Because a write-through cache can be quite effectively implemented in a ring-buffer (with reordering within a moving window for efficiency), you get perfect wear levelling without any complex controller logic. That means that it will work for writing 256MB times the number of rewrite cycles. Cheap flash has 10,000 rewrite cycles. My current laptop has been on for 30 days and has written 172.85GB to disk in this time. That gives 5.76GB/day of writing, or 23 complete write through the cache per day ignore, for now, that some of those were large linear writes, which would probably want to bypass the cache). For 10,000 rewrite cycles, with this usage pattern, it would take 435 day (1.2 years) to wear out the flash. This is, as I mentioned, assuming very cheap flash. Slightly more expensive stuff can get 100,000 rewrites, giving 12 years. If the mechanical parts of a laptop hard drive lasted 12 years, I would be very impressed. They should last longer with this kind of system, because it can batch writes a lot, and reduce the frequency of spinning the drive up and down. You also won't need to spin up the drive to read back data that you've only just written, which could help some poorly performing swapping algorithms (i.e. all of the ones used by 'modern' operating systems).

By the way, flash has a slight weird characteristic that you can write to it with a byte granularity, but only erase it with a block granularity, and it's the number of erases that cause the problems.

Re:Obligatory

[Cyberax]

Nah, it's not worth it. Turning off last access time (BTW, it's turned off by default in Vista :) ) cuts my C++ project building time by 30%. I don't think any kind of intelligent defragmentation will be better.

Re:Obligatory

[Glonoinha]

Having spent hours, days, years studying the effects of hard drive defragmentation, let me put the kibosh on 'intelligent defragmentation' here and now.
Defragmenting the files themselves gives about 20% of the potential benefits of defragmentation.
Defragmenting the file allocation table (FAT on FAT/FAT32 file systems, or MFT on NTFS file systems) gives the remaining 80% of the performance boost potentially given by defragging.

In the big scheme of things, it honestly doesn't matter whether the most recently used files are at the beginning of the drive, next to each other, or on opposite sides of the drive - if the file allocation table (or MFT) is sufficiently fragmented. Frag out the FAT/MFT bad enough over time, and simply defragging the MFT/FAT will make your computer run an order of magnitude faster.

Want the bad news? Windows doesn't ship with a FAT/MFT defragger (well through XP. Not sure about Vista.)
Only way I know to do it is with aftermarket software like Diskeeper (excellent product, BTW, 99% of the time.)

Re:Obligatory

[Glonoinha]

BTW - thanks for sharing the last access update regedit hack. Got any other 30% performance gain tips you want to share, because I will take as many of those as you're willing to share. If I hadn't just replied, I'd be ponying up a mod point or two.

Re:Obligatory

[arth1]

The idea isn't to place the most recently used files where they can be accessed fastest, but to place the LEAST recently used files together, in contiguous areas. Cause chances are that the least recently used files are going to stay the same, and thus not have to be moved around again during a defragmentation. This speeds up repeat defragmentations quite a bit, even if it's a month or two between each time you do it. And when it takes less time, there's a higher chance you'll actually do it.

Re:Obligatory

[Iron Condor]

He said he uses Windows. He didn's say he uses XP embedded (as your link appears to point to an article strictly about EWF which appears to be a feature of the embedded version of XP SP2).

Re:Obligatory

[tomhudson]

"Have 'they' solved the problem of the limited number of writes a flash device device can handle. If it's only going to last a few months and then wear out I won't consider it! Pity the poor fool that forgets to turn off atime updates. "

First, let them solve the number of writes their regular hard disks can handle - 4 drives, bought at 2 different retailers, all dead within 24 hours. It appears that the old Maxtor plant in China that Seagate acquired has some quality-control issues. Best to avoid any 320 gig Seagate drive with a serial # starting with "60F1"

While we're at it, what is it with Seagate only wanting to give a "refurbished" drive to replace a brand-new DOA anyway? This "policy" sucks.. No other business would get away with such tactics.

"Sorry you new car is a lemon. Here, take this used, refurbished car instead." "Sorry the heel broke off your new shoes. Here, have this pair of used, refurbished shoes." "Sorry we're late serving you - we dropped your steak on the floor in the kitchen. But its okay, we refurbished it..."

Re:Obligatory

[Cyberax]

It works on any Windows since Win2k.

You can also use fsutil utility to do the same thing: http://www.microsoft.com/resources/documentation/windows/xp/all/proddocs/en-us/fsutil_behavior.mspx?mfr=true

Re:Obligatory

[Anonymous Coward]

Something similar happened to me a while back. Check your Power Supply and/or buy a lottery ticket.

Re:Obligatory

[tomhudson]

Its not the power supply - the Western Digital on the same bus is still working fine, and its got 9,900 hours on it. How do you think I'm posting?

The drives all have an enormous amount of read/write errors (verified using smartclt to read the SMART data) after a days' operation, and two of them now make that "zing zing zing" noise that heads make when they repeatedly fail during a seek and reset themselves. These drives - ST3320620A - 320 gig, 16 meg cache - are absolutely awful in quality. The two pairs of drives were bought from different companies, in different cities, so its not a "shipping/handling" issue.

I don't want "refurbished" drives to replace 4 brand-new drives that are absolute crap. And their assurance that they "meet the same quality standards as the new drives" just makes me want to laugh. What, they'll also break after a day? No thanks.

I bought these instead of WD drives because the 5-year warranty seemed to be an indicator of their confidence in their product. Instead, its more like the muffler shop scam - you can guarantee a muffler for the life of the car because most people who replace their muffler will either sell the car within the next few years, or it will go to the scrap yard.

The ball is in Seagate's court at this point, but I'm not optimistic. All the recent ST3320620As made in the China factory should be recalled.

Re:Obligatory

[Anonymous Coward]

Have they solved the problem of the limited lifetime of a hard disk?

Re:Obligatory

[ejdmoo]

Vista has these disabled by default (go ahead, look). :)

Re:Obligatory

[Reziac]

About refurbs for warranty replacement -- W.D. does that now too. (Or did, last time I had to RMA a drive -- in 2002. The refurb replacement has been running ever since.)

Presumably they've all gone to this policy because they've learned that most RMA'd drives are not actually defective -- in fact Seagate found that was the case 70% of the time!!

[Indeed, there was a huge rash of "bad" HDs when 40GB came out -- because they were typically formatted as One Big Partition, and FAT32 has a data wrapping bug that can eat data if the partition is larger than 32GB. You can see the problem!!]

That said, IMO if it's bad right away, or dies within an unreasonably short period of use, it's not fair to replace it with a refurb, if only because you just paid for a new drive.

Re:Obligatory

[tomhudson]

That's just it. None of these drives ever worked properly. They all started accumulating errors right away, and making funny noises within 24 hours. They're close enough in serial number that they're probably all the same batch, even though I bought them from 2 different companies located in 2 different cities.

The first two, which I bought 2 weeks ago, never built a proper raid - one drive was always marked as bad-rebuilding, then once it was rebuilt, immediately marked as bad-rebuilding again. The second two, which I bought Saturday, actually succeeded in building a proper raid1 - but then drive 0 started making noises Sunday morning, just like the first batch, and that was it for that raid as well.

Its not like a software RAID1 is all that complicated. Stick the drives on 2 different controllers, mark them as raid, create a raid1 device from them, format, and you're done. Except that, in both these cases, its more a case of "stick a fork in it" type of "you're done."

All the drives report more than 100 million read errors - that's unacceptable for brand new drives. My bet is contaminated media.

Re:Obligatory

[tomhudson]

Wouldn't it be more dependable and a lot quicker to just copy the files (not mirroring the disk - that would still leave the files fragmented) to another drive, then swapping drives so you 're now using a drive that's 100% defragmented, but still have the old drive as your most recent backup?

Or even just use two equal-sized partitions, and swap mount points?

Re:Obligatory

[Reziac]

I'd bet you're right on the contaminated media. Either that, or bogus read/write heads. Sounds like you were getting effectively one sector = one error. Would be interesting to have some data recovery cleanroom outfit give 'em the once-over.

Even something as simple as unfiltered polluted (smokey) air would be the death of 'em, given the literally microscopic tolerances in today's HDs.

I've been specifically warned that *software* RAID is a disk-corruption disaster waiting to happen, and that no wholly-safe software RAID exists. I don't know how true this is, but I don't plan to find out from personal experience. :)

Re:Obligatory

[tomhudson]

I live in a smoke-free environment, so I expect my drives to last a reasonable length of time; in addition, my cases always have extra ventilation. This particular case not only has extra fans blowing air in, but several open bay slots to let hot air out. The WD drive in there is barely warm to the touch right now, after hours of operation (I just stuck my fingers in to check :-).

As for software RAID, all raid is ultimately done in software - either by your operating system, or the firmware on your raid cards' board. With software raid, I can replicate a broken system on another machine without too many problems, and try to recover the data. With hardware raid, if the card is gone, its gone unless I can find another, identical card. Or I can try to recover it by recovering the raw data off the drives, then trying to combine the stripes (if RAID5) on a sector-by-sector basis, or guessing as to which is the bad stripe (2-disk RAID1). With a 3-disk raid1, its a lot easier - go with the majority vote on each sector - but who runs 3-disk RAID1? RAID0? Lost data is lost data, whether you use a software or hardware solution. Sector-by-sector copy to new drives, and pray.

Unfortunately, nothing is going to protect you from that one-in-a-trillion cosmic ray that flips several bits in just the right sequence (though error-correcting ram will lower the odds).

Re:Obligatory

[Silver Gryphon]

Keep every disk less than 70% full and you'll cut way down on fragmentation. Go above 85% full even once, and Windows will hold a grudge. Diskeeper helps calm it down.

Get two or more drives (budget permitting). Partition drives into fast/slow areas. Thousands of small files (random seeks) are a lot faster on the first partition of the second drive (i.e. temp files incl. web browser cache). For Windows, set up the TEMP and TMP environment variables to that partition. I usually do 20-50% for random seeks on the outside of the platter and the rest for larger files that don't seek a lot.

Split the swap across all drives, and increase RAM out the wazoo; file cache is always good. Windows still uses the swap file even with 1GB free... I don't understand why, but the guys in Redmond thought it was a good idea.
Caveat: Move stuff back to the primary drive if you ever need to remove that second drive, or you may get boot time errors.

Pay attention to your own usage patterns (i.e. random or sequential)... if your disk is busy a lot, consider moving an app or two to another physical disk. Development environments can be heavy on random seeks.

I run Visual Studio 2005 Team System at my office; load times with a two-drive setup like that are about 3x faster than a one-drive setup. If your work is video editing or large sequential reads, a RAID setup might be better.

Don't upgrade a 4-year old machine to Vista. If you want Vista, just buy a new one.

Re:Obligatory

[Junks Jerzey]

Anyone know how to disable file last access timestamps under OS X?

Re:Obligatory

[Anonymous Coward]

I don't know if someone convinced you of this or you're just making this up, because it's the biggest load of bs I've ever heard.

Re:Obligatory

[Anonymous Coward]

Free space consolidation is also an important piece, especially when writing. PerfectDisk does this, along with all system metadata files, which Diskeeper does not address (other than MFT).

Joe Abusamra
Raxco Software

Re:Obligatory

[LorentzForce]

With the automatic background defragmentation an option these days, remembering to defrag or motivating yourself to defrag (lol!) is pretty much a non-issue :) . Automatic defrag is a great solution for those who don't want to bother with scheduling a defrag or manually executing one the old fashioned way. It saves time and manpower if the server + workstations can largely take care of themselves atleast for the defrag side of things.

Couldn't this be done in software?

[morgan_greywolf]

Like with a USB flash drive and a regular SATA HDD? Maybe an application that sits in the background, mirroring the contents of the flash drive to the HDD?

Re:Couldn't this be done in software?

[PalmKiller]

Yes, but it would be much slower and would use cpu cycles to do it.

Re:Couldn't this be done in software?

[morgan_greywolf]

But with USB 2.0, still potentially faster than a bare HDD, right?

Re:Couldn't this be done in software?

[AmericanInKiev]

Integration can solve a key point which is data integrity during an abrupt power event.
(see above).
AIK

Re:Couldn't this be done in software?

[TheRaven64]

Depends what 'this' is. The most interesting use for a hybrid drive is using the flash as a write cache. You stream write requests into a ring buffer in the flash, and it processes and retires them, skipping ones where you have already written the block once and reordering within a moving window of 1-2MB. This gives you the following advantages:

• Writes will complete when you power on if the power failed in the middle.
• You don't need to spin up the hard drive for small writes, even if they are important.
• You can quickly read back recent writes from the cache (useful poorly performing VM subsystems).

Re:Couldn't this be done in software?

[Goaway]

Yes. Vista supports this.

Slashdotted...

[andrewd18]

The server at www.pclaunches.com is taking too long to respond.

I think it's pretty safe to say that pclaunches.com isn't using hybrid hard drives.

Re:Slashdotted...

[superpulpsicle]

LOL

Re:Slashdotted...

[arth1]

I think it's pretty safe to say that pclaunches.com isn't using hybrid hard drives.

Um, no, that's not pretty safe to say. Thing is that the hybrid drives have not shown any speed benefits in real world usage tests, and have often been slower than comparable sized drives.

Re:Slashdotted...

[Anonymous Coward]

Sod off, you humorless pedant.

Re:Slashdotted...

[Wog]

(Score: -1, missing the joke)

Re:if it's a hybrid

[Anonymous Coward]

Furthermore, if they're going to have so little flash memory, why not have 256 MBs of cache instead... Seeing flash memory wears out and is more expensive.

IDE?

[DogDude]

Anybody know if it's IDE or SATA? There was an article a while back about Seagate dropping IDE by the end of this year. At least for me, it's a brick if it's not IDE.

Re:IDE?

[systemic chaos]

Anybody know if it's IDE or SATA? There was an article a while back about Seagate dropping IDE by the end of this year. At least for me, it's a brick if it's not IDE. Amen. And where's my 16-bit version of Vista on a nice solid stack of floppies?

Re:IDE?

[Reziac]

You can get SATA to/from IDE adapters -- look for 'em cheap at cablenbits.com and tekgems.com (I've dealt happily with both companies).

But I don't know how well the adapters work, and I can tell you finding a PCI SATA adapter that handles over 400GB, and doesn't cost more than a new motherboard, is tough to do!!

Re:IDE?

[Anonymous Coward]

You'll be missed!

--Randal

Re:IDE?

[Anonymous Coward]

Do not follow this posters' advice. DO NOT buy one of those adapters. They are known to be faulty in numerous respects to the SATA protocol, and have been reported (by both Linux and FreeBSD users) to cause data corruption or data loss, as well as DMA timeouts and other anomalies.

Buy yourself a cheap SATA controller card if you lack SATA ports. Buy yourself a PATA controller card if you lack PATA ports.

Re:IDE?

[Reziac]

An AC still modded at zero reports, "DO NOT buy one of those adapters. They are known to be faulty in numerous respects to the SATA protocol, and have been reported (by both Linux and FreeBSD users) to cause data corruption or data loss, as well as DMA timeouts and other anomalies. Buy yourself a cheap SATA controller card if you lack SATA ports. Buy yourself a PATA controller card if you lack PATA ports."

That's very interesting; anyone else have information on this??

The problem is, there is no such thing as a cheap SATA2 controller that handles 500GB drives. Having searched extensively, the lowest price I've seen was $85. Not exactly "cheap". The $35-and-less cards max out at 300 to 400 GB. Furthermore, it seems only SATA1 comes in PCI; SATA2 adapters apparently only come in PCIx, until you get into the very pricey ones.

Re:IDE?

[Craig Davison]

The Promise SATA 300 TX4 is a PCI SATA2 controller with 4 ports. But like you said, it's $80 (in the store down the street. Probably cheaper online).

BTW don't use this card in Linux. The drives are ordered backwards (so forget about booting with it), and the performance sucks. Seems fine in Windows.

Re:IDE?

[Reziac]

That's the one and only. $80 is the best price I've seen. :( For that you might as well buy a new motherboard with the required ports. (Tho I've noticed prices on good motherboards are going up, up, up...)

Interesting about the drive order, what's up with that?? or just no good linux drivers for it yet?

Yeah except I prefer speed over power saving

[beckerist]

http://www.pcworld.com/zoom?id=138102&page=1&type=table&zoomIdx=2 -attached to- http://www.pcworld.com/article/id,138102-c,harddrives/article.html
Both hybrids, Samsung AND Seagate were not only more expensive, they were considerably slower in tests vs. a traditional harddrive. I understand the drive to be green, but I think I'm going to wait a few years before jumping on this bandwagon!

Re:Yeah except I prefer speed over power saving

[beckerist]

PLUS, they both require Vista for full functionality....ahem...NO THANKS, XP works just fine for me!

--beckerist

Re:Yeah except I prefer speed over power saving

[damburger]

Unless you are running an internet server where the access speed is going to be dwarfed by the latency, and the extra cost of the more energy efficient drive is going to pay for itself in a couple of weeks of heavy use.

Re:Yeah except I prefer speed over power saving

[skiflyer]

They tested two 5400RPM hybrid drives against a 7200RPM standard drive... the results are as expected. Now I do agree that if I'm going to shell out the extra cash for the hybrid I probably want the 7200 drive too... and I definitely agree that I'd wait a couple generations (dunno about years)

Re:Yeah except I prefer speed over power saving

[Anonymous Coward]

You do realize the "normal" HD they used was a 7200 rpm one, right? 2 5400 RPM hybrids vs. a 7200 rpm disk, in streaming disk tests...gee, I _wonder_ which one will be faster...

Re:Yeah except I prefer speed over power saving

[Orange Crush]

It's my understanding that these drives are being targeted at laptops to improve battery life at a reasonable cost premium for what you get.

Re:Yeah except I prefer speed over power saving

[StikyPad]

That's because flash is extremely slow for writes.. somewhere in the neighborhood of 4MB/s. I've personally never seen faster in my testing, even so-called "60x" or "96x" chips. Better flash may well exist, but I doubt they're using anything faster in commodity hard drives.

As for the power savings, I imagine the bulk of the savings comes from using non-volatile RAM, so constant power isn't required for the cache. A few extraneous spin-ups *might* be avoided, but you don't really want to keep writes cached any longer than necessary. Maybe the savings are significant, but even if it's 20% more efficient, and the HD accounts for, say, 20% of your power consumption (which seems to be a generous estimate), that's only a 4% savings overall. It may useful in conjunction with other power-saving methods perhaps, but seems fairly useless on its own. Perhaps worse than useless, if performance is at all important to you.

Re:Yeah except I prefer speed over power saving

[Anonymous Coward]

XP works just fine for me!

Good one! Too bad the mods have no sense of humor.

Re:Yeah except I prefer speed over power saving

[drix]

I understand the drive to be green, but I think I'm going to wait a few years before jumping on this bandwagon! In a few thousands years, when anthropologists are piecing together the clues to the decline of our civilization, I really hope they come across quotes like this.

Needed for Vista?

[QCompson]

Good. These are required in order to run Vista. Or wait...

You mean like ...

[ThirdPrize]

this?

Re:You mean like ...

[morgan_greywolf]

Um, yes. I guess I initially misunderstood ReadyBoost, thinking it was a way to add 'RAM' to a system using a USB flash drive. Anyone know of a similar tool for Linux?

Re:You mean like ...

[TopSpin]

The link you provide leads to this explanation:

When you're not actively using your computer, background tasks--including automatic backup programs and antivirus scans--run when they will least disturb you. These background tasks can take up system memory space that your programs had been using. On Windows XP-based PCs, this can slow progress to a crawl when you attempt to resume work. Low priority background processes are flushing pages needed by foreground processes which leads to a lengthy series of page loads when you resume work with those processes. Solution: hide some memory on the storage device where the pager can't screw it up. Brilliant.

Disclaimer: this problem is most emphatically NOT unique to Microsoft platforms.

The grandparent's question is a good one. I think the answer is 'yes', but it would require virtual memory management that incorporated far more knowledge about usage patterns of pages. Apparently such things are intolerably ugly to those who engineer virtual memory managers.

Disclaimer 2: IANAKD

Re:You mean like ...

[polymath69]

Easy.

Say you have a memory stick installed and recognized as /dev/sda. Use cfdisk to create partition 1, type 82. mkswap /dev/sda1, then swapon -p 100 /dev/sda1.

You could also use a file within a file system on the memory stick, with mkfile and mkswap followed by swapon, but that would have higher overhead so lower performance.

Re:You mean like ...

[morgan_greywolf]

Right. That's like setting up swap on USB disk. But that's not what I was talking about.

Re:You mean like ...

[polymath69]

I guess I must be unclear on what ReadyBoost is, then. From MS's page it sure sounds like "priority swap to USB", only they try to couche the whole thing in ambiguity. What is it, then, if not this?

Re:You mean like ...

[drsmithy]

I guess I must be unclear on what ReadyBoost is, then. From MS's page it sure sounds like "priority swap to USB", only they try to couche the whole thing in ambiguity. What is it, then, if not this?

Essentially, it's a DIY hybrid hard disk - take normal hard disk, add thumbdrive and ReadyBoost -> most of the benefits of a hybrid drive. It has nothing to do with swap (apart from possibly improving swapping performance by caching).

With that said, the ReadyBoost web page linked above doesn't explain this very well - but that's probably because it's trying to present what is a fairly complicated concept to non-technical users. The Wikipedia page is better and there's almost certainly detailed information in MSDN somewhere.

Re:You mean like ...

[Anonymous Coward]

From my understanding, the major difference is that what's swapped to ReadyBoost is also written to HDD; it can be paged back in from ReadyBoost faster than in from HDD. (A page isn't considered swapped out until it's written to both devices.)

The theory being is that if the memory stick is pulled out, you don't have a system crash on your hands (just falls back to paging in from conventional disc).

(Other major difference is that it's encrypted. There's no need for a key managment system 'cause the randomly generated key is merely kept in real RAM. Yes, I realise Linux can do encrypted swap; but not by default. )

http://blogs.msdn.com/tomarcher/archive/2006/06/02/615199.aspx

Re:You mean like ...

[Goaway]

I guess I must be unclear on what ReadyBoost is, then. That hasn't stopped a thousand slashdotters before you from claiming that Linux can totally do it too, either, so don't feel too bad about it.

NCQ

[ficken]

I guess they just upped the 16MB cache many drives have now. Makes sense to do with native command queuing being as versatile as it is.

Pointlessly small...

[Joce640k]

I can buy a 2Gb USB pen in the shops for $15 so why put so little flash memory in it?

I'd be happy pay another $10 for a decent amount.

Re:Pointlessly small...

[Anonymous Coward]

Are you so sure that the flash memory in both is the same speed?

Re:Pointlessly small...

[toddestan]

Are you so sure that the flash memory in both is the same speed?

Not to mention the same quality. A USB pendrive could get by with the cheap stuff that only has 10,000 writes. I would guess that for a caching operation, Seagate would have to go with the more expensive stuff that gets you 100,000 or 1 million writes.

Half a Solution

[Nom du Keyboard]

Seems to me that writing to flash is rather slow, Rereading the data afterwards would be much quicker, meaning your best approach would be to find a way to put write-once-read-many data there. What OS supports this?

Re:Half a Solution

[g0at]

Seems to me that writing to flash is rather slow, Rereading the data afterwards would be much quicker, meaning your best approach would be to find a way to put write-once-read-many data there. What OS supports this? Every OS known to man; it's called SDRAM.

I mean seriously, if the OS knows that there are a few hundred GB of data to which it regularly needs access, why not cache them in regular system memory?

b

Re:Half a Solution

[Anonymous Coward]

How much do you pay for that computer with a few hundred GB of system memory?

Re:Half a Solution

[g0at]

Crap, oops; I meant a few hundred MB.

-b

Re:Half a Solution

[Anonymous Coward]

If it were nonvolatile, I'd buy one this second.

256MB is more than enough space for the Linux kernel, bootloader and whatever goodies you needed to get your box revved up in a snap. Apps and everything not required for boot goes to the hard drive.

Re:Half a Solution

[Glonoinha]

A few hundred Gigs of system memory ought to be enough for anybody.

Article is /.'ed

[Seakip18]

Can't read the article but this will help understand about the Hybrid drives.
Since laptops can't support the faster speeds that their desktop brethren, any access time improvement is desirable. You can keep your most frequently used data on the Flash or as a buffer, such as during a movie. Another benefit is that flash takes less energy to read than a HDD.
Here's also a review of the drive itself

I'd rather it allowed the drive to spin down.

[Joce640k]

Windows has this thing to let the drive go to sleep when you're not using it... ...except it never does because Windows is always syncing it or doing something. It never gets enough idle time to actually spin down.

If these drives could fool Windows into letting them go to sleep we might be onto something.

Re:I'd rather it allowed the drive to spin down.

[Reziac]

Better not tell my yonder XP box, then. Its HDs have been asleep all day, the lazy things.

But I turn off indexing service, which doubtless makes a big difference.

Re:I'd rather it allowed the drive to spin down.

[Trogre]

I'm not sure that would be good. If you believe Google's findings, you'll kill your hard drive's bearings doing that sort of thing.

Re:I'd rather it allowed the drive to spin down.

[toddestan]

I'm not sure that would be good. If you believe Google's findings, you'll kill your hard drive's bearings doing that sort of thing.

If you go back and (re)read Google's study, you'll find that Google has little to say about power cycling harddrives, as they let them run continously. While Google did note a weak correlation, they speculated that this might be caused by already problematic machines that needed to be powered down and repaired more often.

I agree that convential wisdom does say that lots of power cycles is bad though.

Re:I'd rather it allowed the drive to spin down.

[backwardMechanic]

Battery life or bearings - you choose.

Re:I'd rather it allowed the drive to spin down.

[midnighttoadstool]

Actually Windows is quite well behaved in this respect.

Try running a harddrive activity utility to track down the app accessing your platters. It will probably turn out to be 3rd party. Process Monitor should do it.

Re:I'd rather it allowed the drive to spin down.

[cecil_turtle]

I think the grandparent poster was speaking more specifically of Vista for hard drive activity (I'm guessing), I don't know if you've tried it out but the hard drive in Vista just grinds away continuously and you have no idea what it's doing. Especially if the computer was turned off for any period of time - if it sits for a week suddenly Vista thinks it has all kinds of maintenance tasks to catch up on even though nothing changed.

Re:I'd rather it allowed the drive to spin down.

[Reziac]

Ah, Vista. Haven't tried it yet myself... Gods know what it's doing... hell, maybe it's just jacking off!

XP also has housekeeping fits at odd times, tho it doesn't usually wake itself from a sound sleep to do so. I suppose Vista's logic was "do housekeeping while the computer is otherwise-idle" but got a little carried away on their definition of "idle".

System Restore likes to have a cleaning fit when XP first wakes up, and I imagine it and other janitors do much the same on Vista, only more of it (Vista having so much more bloat).

I'm still waiting...

[ivormi]

For Hybrid Hard Drives to live up to promises. After a bit more digging - There is still a lack of results from this drive, although boot time and power savings are starting to show up. RAM caches have been around for years, and getting even 1 GB of flash memory is getting down to pretty reasonable levels. Why is this commanding a 30% premium and delivering unspectacular benefits? Unless there's a solid standard behind addressing for HHD's exists, there's no point in blaming BIOS or Vista for a problem that could also be addressed in on-drive logic.
Meh.

I'm not an expert on flash media

[Opportunist]

But don't they die after a few thousand write cycles? Anyone with some information about this matter able to provide some insight?

Re:I'm not an expert on flash media

[The Living Fractal]

I read an article somewhere that showed how a flash based drive could outlast a platter drive by efficient use of an algorithm that rotated through the bits. I don't recall any further information on this though, such as performance impact. Sorry for the lack of a link. I am sure you can google it though. :)

Re:I'm not an expert on flash media

[Red Flayer]

Sorry for the lack of a link. I am sure you can google it though. :)

Aww, come on... this is Slashdot.

Most of us can be barely bothered to read the summary, let alone TFA... and you want us to google for a link?

Might as well ask for a never-ending supply of beer*, and 12 nekkid virgins to be awaiting your return home after work tonight to satisfy your every whim.

*or Mountain Dew, depending on your age/preference.

Re:I'm not an expert on flash media

[walkie]

http://en.wikipedia.org/wiki/Wear_levelling

Re:I'm not an expert on flash media

[Anonymous Coward]

This comes up every...single...time there is a flash article on Slashdot. Flash supports millions of write cycles, and the controllers these days automatically level the writes across random sectors (since any one is as fast as any other). MTBF is higher than standard platter-based HDs.

Go to Seagate's website

[flatulus]

http://www.seagate.com/ has a press release on their home page.

Hybrid Irony

[Mingco]

It's ironic that hybrid cars save energy by spinning a platter and hybrid hard drives save energy by not spinning a platter. It's like blowing on your coffee to cool it and blowing on your hands on a brisk day to warm them. If we could just hook these devices up in round-robin, we'd have a perpetual energy machine!

Re:Hybrid Irony

[jandrese]

By "spinning a platter" you mean "has an electric motor"? That's the only platter like object I can think of in a Hybrid drivetrain, and unless you're using some very strange motors they aren't platter-like at all.

In short: What are you talking about?

Even shorter: Are you high?

Re:Hybrid Irony

[19thNervousBreakdown]

He's thinking of flywheels as platters.

Re:Hybrid Irony

[Anonymous Coward]

What charges the motor -- the combustion driven rotation, which is channeled through the flywheel.

Maybe you're high, because I don't see how someone could miss this if they know even a tiny bit about automobiles of any kind.

Re:Hybrid Irony

[TheClam]

-1, Alanis (not irony)

Re:Hybrid Irony

[jandrese]

There's no special flywheel in the Hybrid though, just what you might find in a conventional gasoline engine. From what I've seen, the Prius (which, being the most popular I have the most experience with) doesn't have a flywheel at all on the drivetrain.

Re:Hybrid Irony

[Spacezilla]

irony
  -adjective
  consisting of, containing, or resembling iron.

This isn't about being green

[Sycraft-fu]

This is about having longer laptop battery life. These days, processors are pretty good at throttling back. So the next big consumers are the harddrive and the screen (or rather its backlight). Well, hybrid harddrives offer a potential solution. Cache frequently needed data and small writes to flash, and you can spin up the drive platters less often. That saves power which increases the time you get on battery. Also it actually will make a laptop MORE responsive in that if the disk is spun down, the flash can handle things as it spins up so everything doesn't have to come to a halt waiting for it.

I don't know how much of a use these will be in desktops, but in laptops it seems like a really good idea. Also, Seagate drives normally perform slower than the competition. In basically all the tests I've seen, their drives are on the bottom. Of course we are talking a difference of a few percent at most, and perhaps that's also the reason their drives last longer. Maybe they don't push them so hard.

256 MB?

[Azuma Hazuki]

What's going to go in that 256 MB though? No modern OS can fit in there except a stripped down Linux or BSD, and if this is for mainstream use most people won't be running those. Using it as swap space or /tmp would degrade it quickly even with the newer, longer-lasting materials. So what can this be used for?

Re:256 MB?

[schnikies79]

Storing hibernation data/standby data, hence why it's for notebooks.

Like the Transistorized Vacuum Tube Radios?

[Cliff Stoll]

Around 1956, electronics makers began selling hybrid radios with both vacuum tubes and transistors. Emerson's vest-pocket portable model 843 used tubes in the rf stages and a pair of plug-in transistors for audio output. A 6 volt battery lit up the tubes and transistors, while a 67 volt battery kept the tubes' electrons jumping from cathodes to plates.

From Emerson's adverts: "Transistors are so tiny they must be seen to be believed. Transistors are so sturdy they won't break... They will last for life!" and give "greater power without distortion - full reproduction of voice and instruments, balanced tone quality, and greater power output with less distortion, not to mention low battery drain"

What other mixed hybrids have came along? Was there ever a hybrid horse and car?

Re:Like the Transistorized Vacuum Tube Radios?

[newr00tic]

Hi, Cliffie.

There's a hybrid horse on the market; -it's called a mule.. ;)

(The Amish, if noone else, are probably reluctant to make use of them, along the same lines as slashdotters despise Vista, I'd imagine.)

Re:Like the Transistorized Vacuum Tube Radios?

[jbeaupre]

You mean the horse buggy?

Side note: a buddy of mine owns the second Oldsmobile ever made. It actually has the first Oldsmobile engine (the guys used engine #1 when they built car #2). Anyway, this car is almost identical looking to a horse buggy, with just a few modifications. The wheels are driven by a belt from the engine hidden in the back. So in effect you had a hybrid horse-buggy/automobile (ok, with no horse)

Re:Like the Transistorized Vacuum Tube Radios?

[Mursk]

Go Horsey Horseless!

http://www.time.com/time/specials/2007/article/0,28804,1658545_1657686,00.html

Re:Like the Transistorized Vacuum Tube Radios?

[langelgjm]

What other mixed hybrids have came along? Was there ever a hybrid horse and car?

You bet.

Re:Like the Transistorized Vacuum Tube Radios?

[Tweekster]

Back to the Future 3

Re:Like the Transistorized Vacuum Tube Radios?

[Fear the Clam]

Was there ever a hybrid horse and car?

No, but there was a hybrid horse and rider.

Re:Like the Transistorized Vacuum Tube Radios?

[Agripa]

To be fair, early transistors inexpensive enough for tube replacement would not have the performances necessary for RF and IF applications. Hybrid designs were quite common until it became cheaper to use transistors throughout.

Re:Like the Transistorized Vacuum Tube Radios?

[hcdejong]

By the end of WW2, aircraft engine technology was transitioning from pistons to turbines. The last generation of piston engines relied heavily on turbochargers and/or superchargers. Engines like the Napier Nomad and the Wright R3350 turbocompound can be considered hybrids: some of their output power comes from the piston engine, but some comes directly from the turbines.

Re:Like the Transistorized Vacuum Tube Radios?

[Anonymous Coward]

Steam and sail hybrid ships were quite common for a while.

Re:Like the Transistorized Vacuum Tube Radios?

[smellsofbikes]

I was thinking about turbocompounding, too.
Also the B-36 with a mix of jet and prop engines.
Or, in my own field, switching power regulators that are complex IC's, some of which have the switching FET internal to the chip and some of which use external FET's -- basically the same thing as the tube->transistor evolution, one more step along, as we go from discrete components to a mix of discrete and IC.
Likewise, there are lots of cars out there with incandescent-glowing-wire lightbulbs inside the car, halogen-arc-discharge headlights, and LED taillights. We're pretty clearly moving to all-LED lighting, but we're not there yet. Or lasers, which traditionally used flashlamps or arcs to power them, but are slowly moving to using laser diodes for pumping, and presumably will end up being just laser diodes.

Re:Like the Transistorized Vacuum Tube Radios?

[Anonymous Coward]

sort of: 1899 Horsey Horseless

Vista requirement

[ehiris]

Does that mean that the drives will not work with Linux?

Re:Vista requirement

[mha]

No they will, you just cannot use ReadyDrive technology. no big deal, since Vista itself can't make use of it (see my other response). Basically, that drive doesn't deserve a ./ headline at all, nothing new, nothing to be seen here, don't assemble on the Internet, go home and let the authorities take care of this... :-)

Made for Vista ReadyDrive - which is USELESS

[mha]

What is ReadyDrive:
http://www.microsoft.com/windows/products/windowsvista/features/details/performance.mspx

I'm summarizing what I learned from the German c't computer magazine, which has tested the various new technologies like ReadyDrive and others in Vista and also tested Flashdrives and Flash memory in general. Read the current issue of this magazine for in-depth analysis.

1) Pure Flash disks have only ONE advantage over harddisks: they are less sensitive to mechanical stress. In real-life scenarios, they don't safe power, and they are most definitely not faster than 2.5 inch drives. They ARE faster than 1.8 inch ones often used in ultra-mobile PCs, so there they indeed provide a benefit. For everyone else: especially write performance sucks compared to modern 2.5 inch disks, and read performance is at most en par. True, they don't need to position any heads so random access should save time - but according to the real-world tests made by c't that benefit isn't noticeable.

2) c't testers were very suspicious about how long Flash memory could survive as HD replacement where writing happens all the time, and yes, Flash cells have a limited lifetime, one cannot write too often. That's the theory. In practice c't testers were unable to make even the cheapest Flash USB stick show any sign of memory loss even after something like 16 million write cycles, when they gave up further testing because that's many many years of real-work usage. (pg. 104 of c't 21/2007)

3) Intel TurboMemory or MS Vista SuperFetch, ReadyBoost or ReadyDrive were shown to provide no measurable benefit AT ALL.

Suspicion of Hitachi and others seems to be that the current implementation in Vista isn't quite finished and SP1 should provide an update, and second the amount of Flash memory is waaaaaay too small.

Original article (German): http://www.heise.de/ct/07/21/100/

Re:Made for Vista ReadyDrive - which is USELESS

[mha]

I think I should mention that pure Flash disks (SSDs) have real advantages are are very close, unlike those Hybrids for which no real purpose can be seen at this point. SSDs, if available more cheaply, are interesting not just for some Notebooks but also for servers - they need much less power... am I contradicting myself? no, looking at just the disk that's true, but c't measured real-world performance, and if you get chipset graphics instead of Nvidia in your notebook you safe MUCH more than by inserting an SSD instead of an HD. In c't tests the difference was a few minutes of on-time in a notebook. Also, in servers you have fast power-consuming disks and not notebook disks so the saving is much higher. In a rack with hundreds of servers it's A LOT. Another aspect: for server disks random access times are much more important than for Notebooks/Desktop PCs. So right now the main beneficiary of SSDs would be - servers, not Notebooks! Remains the price issue...

But again, the disk this headline is about is none of the above. It is just useless.

Re:Made for Vista ReadyDrive - which is USELESS

[glwtta]

In real-life scenarios, they don't safe power, and they are most definitely not faster than 2.5 inch drives.

I'm somewhat skeptical about that one - is there a non-German source for those tests?

Re:Made for Vista ReadyDrive - which is USELESS

[mha]

See my response to my response, please :-)

Yes of course they safe some, but not enough to matter in a notebook. Especially when you have a notebook with a graphics card instead of chipset graphics the difference is negligible.

Facts from the c't article (great, I get mod points but all I do is translate from German :-) ):
- The Seagate hybrid disk uses 1.1W when doing nothing
- it uses 3.5W when reading/writing
- spikes of >5W can be seen when starting the disk platters

- The tested SSD uses 0.6W when doing nothing
- 0.9W reading/writing

So if the disk would be doing something PERMANENTLY one could save 2.6W.

They tested a Dell D630, which with the CPU doing nothing and the HD not spinning used 11W. The Dell had chipset graphics, the other tested notebook had an Nvidia and used 21W doing nothing (and some notebooks use >30W doing nothing!) - you can already see the disk being MUCH less of an issue! Also, since in a notebook (see my above mentioned response to my own article about server disks) a disk is hardly doing something all the time the gain of using an SSD is negligible.

In a test of the Dell notebook SSD vs. HD they managed to get it to run 40 minutes loger. BUT that was with a test that permanently wrote something (copying 4GB files back and forth). Using a benchmark that tests real-work behavior (office work in this test case, games may differ) the difference in all the many different tests they did was a mere few minutes.

Now go out and learn German and buy that c't ;-)

Re:Made for Vista ReadyDrive - which is USELESS

[sid0]

Could you please post test results? I really can't imagine how SuperFetch is useless.

I'm not so sure about ReadyBoost -- did they test it in the conditions it is designed to handle, i.e. very high loads with lower amounts of RAM? I don't think it'll show improvement under low loads.

Also, given that the first hybrid hard drive was released some hours ago, did they land a "pre-release copy" to test?

Re:Made for Vista ReadyDrive - which is USELESS

[mha]

Yes they did because they couldn't quite believe it was THAT useless themselves...

But I can't post the whole article, which actually is three articles each several pages long (not interrupted by half-page ads as is common in US computer magazines)! There are lots of copyright threads here on this very website lately ;-)

I think I pointed out that the suspicion is not so much that the *concept* is bad, but that the implementation in Vista isn't really finished in their rush to get it out, and that retesting after SP1 is available early next year might show different results.

Re:Made for Vista ReadyDrive - which is USELESS

[mha]

Also, given that the first hybrid hard drive was released some hours ago, did they land a "pre-release copy" to test?

They tested Samsung MH08HHI and MH16HIJ with 256MB Flash memory. The Seagate was not tested. I don't think there will be a significant difference between the two, none of the few remaining disk manufacturers has any significant technological lead over the others I'd venture to say.

For testing SuperFecth they used a robot(!) to move the mouse during testing, which Vista requires because this is how it detects if the user is doing anything, important for SuperFetch optimization.

But SuperFetch has no direct connection with the Flash memory. Only if there is not enough RAM SuperFetch uses the other techniques, especially ReadyBoost. And that does not benefit. So SuperFetch is a good idea, but using Flash memory is not at least right now. Invest the money more wisely by adding RAM. Or wait until they offer Hybrid disks with 5GB Flash and Vista with Service Pack 1.

Re:Made for Vista ReadyDrive - which is USELESS

[sid0]

SuperFetch caches frequently used files into RAM. It doesn't have anything to do with Flash memory. If you open Task Manager in Vista, you'll see something like

Memory
Total: 2047
Cached: 1347
Free: 1

1347 MB of RAM is being used as a cache for files by SuperFetch. Are we talking about the same thing?

Re:Made for Vista ReadyDrive - which is USELESS

[mha]

I know and I never said anything different.

Drives are out, no performance increase....

[Fallen Kell]

With a famous quote, "By the second generation products will see the system benefits", by Melissa Johnson, a product manager at Seagate. http://www.extremetech.com/article2/0,1558,2188425,00.asp?kc=ETRSS02129TX1K0000532 http://www.dailytech.com/article.aspx?newsid=9195

Integrated - NOT!

[CustomDesigned]

Apparently, both the Seagate and Samsung drives are not integrated, since they require Windows Vista to actually used the flash. Seems really stupid to me. The drive ought to just accumulate writes in the flash so as to avoid spinning up the disk until the flash is full. Use regular RAM for read caching. On power failure, accumulated writes are still in flash - unlike with a RAM cache. They talk about faster bootups too, which would require keeping sectors read shortly after powerup in flash until next powerup.

Why does any of this require OS hooks? If you're going to have OS hooks, you might as well glue a USB thumb drive to the hard drive and be done with it. (And in fact, an md-like linux driver to combine two block devices in a manner like the above would be a great hack.)

Re:Integrated - NOT!

[Anonymous Coward]

It requires OS support to tell the drive what to cache. How does the drive know which sectors are used during boot-up? How does it know which sectors are accessed the most? How does the drive decide what fraction of the drive to allocate to boot-up sectors versus most-accessed sectors versus write-back caching?

It is quite possible that without OS support the drive could use the flash as a giant write-back cache. Keep in mind, though, that this would not give you the largest benefits of hybrid configuration, which are faster boot-up and the ability to operate with the drive spun-down.

dom

Re:Integrated - NOT!

[CustomDesigned]

How does the drive know which sectors are used during boot-up?

Heh. I know reading comprehension is not the greatest when trying to sneak in slashdot at work, but I'll repeat the (hardly the only) idea I made previously: Keep the sectors read shortly after the last power up in NV read cache.

How does the drive decide what fraction of the drive to allocate to boot-up sectors versus most-accessed sectors versus write-back caching?

Again, to repeat from my previous post: don't use flash for read cache, except boot up sectors per above. Use volatile RAM for read cache in operation. To elaborate a little, I would record a weight that decays rapidly after power up/reset. Would resetting the drive during operation (perhaps because of some hardware lockup) destroy the boot cache? Absolutely - but it will get refreshed again at the boot - no different than loading a new kernel.

The allocation between write cache and boot cache depends is adjusted at each boot. Sectors that are read again at boot could slide the partition. Note that *which* sectors can be recorded in NV without recording their contents - for the purpose of adjust the boot cache size. Once the OS starts reading sectors that are different between boots, they don't go in the boot cache. Again, the longer the elapsed time since boot, the less important the sector is to the boot cache (since the user is already bored) - hence the decaying weight.

Re:Integrated - NOT!

[DES]

Apparently, both the Seagate and Samsung drives are not integrated, since they require Windows Vista to actually used the flash. Seems really stupid to me. Actually, it's really smart, because the OS knows a lot more about the data it writes to disk than the disk does, and especially about ordering and integrity concerns, and can therefore make more efficient use of the cache than the disk can. For instance, you could build a ZFS pool out of these drives and use the flash to store the intent log.

Re:This is Great -- So what you are saying is...

[rhartness]

"256MB should be enough for anyone." Where have I heard something like that before?

How can it have the speed of a solid state drive?

[Anonymous Coward]

It's clearly not a solid state drive, it just incorporates an extra passage in the flow of data that is "solid state", but the drive is never faster than the weakest link in the chain - and that's the magnetical disc, so how does this drive get its speed? In disc writes, sure, but consequent reads despite the flash buffer cache?

Sounds like a stupid idea that only costs the customers alot of money while still giving them alot less storage.

Actually, laptop HDDs don't consume that much

[vlad_petric]

Most of the power (~90%) is still consumed by the processor and screen. Bottomline - these HDDs are pretty much pointless right now.

surviving falls

[H310iSe]

Surviving falls for a HD is a Very Good Thing. But they claim the drive when spun-down (just using the flash) can survive a 6 foot fall or 900 Gs of deceleration. 900? I'm having a hard time understanding how 6 feet of free fall ends in 900g of force - does that mean, say, a 10 lb object dropped 6 ft will hit the floor with the equiv. of 9,000 lb of force? Anyone who paid attention in HS physics care to elaborate?

I have done some googling but gotten the greater part of nowhere

Re:surviving falls

[Anonymous Coward]

You're comparing forces to accelerations.

If you wanted to calculate the deceleration of an object when it strikes
the ground I suppose you'd need to find the velocity of the object
when it strikes the ground. That's fairly simple using 1 dimensional
equations of kinematics. You'd also need to find the amount of time taken
for the object to go from this velocity to zero. Probably an extremely
small number. Thus achieving a huge deceleration for a minute amount
of time isn't that interesting.

What would probably be more interesting is how long it can survive in
a sustained deceleration. Then again, unless you're doing spaceflight with
these hard drives even that isn't very insightful.

Re:surviving falls

[dwye]

> I'm having a hard time understanding how 6 feet of free fall ends in 900g of force

It's not the fall that kills you; it's the sudden stop at the end.

I assume that the stop after a 6 foot fall generated that much deceleration (for an instant) in their worst-case testing rig, so thereafter they developed to the 900 G figure.

Re:surviving falls

[kasperd]

> I'm having a hard time understanding how 6 feet of free fall ends in 900g of force

It's not the fall that kills you; it's the sudden stop at the end.

Exactly. So the 900g figure simply means it took 900 times as long time to accelrate during the fall than it took to slow down when it hit the floor. Let's assume a fall from that height took 9 seconds (I could have computed the exact value if the height had been given in standard units), then it means it must stop in just 0.01 seconds when it hits the floor to have a 900g force. (Not completley accurate since the force during those 0.01 seconds is probably not constant, but you get the idea.)

Re:surviving falls

[flatulus]

The 900G represents how fast the object was going when it hit the ground, and how fast it stopped going down. The further something falls, the faster its velocity will be at impact. When a laptop hits the ground, it STOPS moving down virtually instantaneously. So from a standpoint of "acceleration" (which is what the G represents), the acceleration is very high -- or put another way, the change in velocity from X m/sec to 0 m/sec happens very very fast, hence the rate of change of velocity (which is what acceleration represents) is very high.

Now, if you want to REALLY think about high acceleration, consider a pool player shooting a target ball with high force. The cue ball contacts the object ball, and due to the virtually zero elasticity of billiard balls, the target ball *accelerates* from zero to its "terminal velocity" (whatever that might happen to be) in an exceedingly short time. So the "G force" of the impact would be correspondingly high.

Re:surviving falls

[bentcd]

The equations of motion can be found at
http://en.wikipedia.org/wiki/Equations_of_motion#Linear_equations_of_motion

In this particular case,
vf = vi + a * t
seems most appropriate (where vf is final velocity, vi is initial velocity, a is accelleration and t is duration of acceleration).
Assuming vf = 0 and solving for a, we get
a = vi/t (1)
Solving for t yields
t = vi/a (2)

"vi" after a 6-foot fall can be determined by using another of the equations:
vf^2 = vi^2 + 2ad
(where d is distance in meters, let's use 2 meters for 6 feet. We must assume that vi is zero.)
vf^2 = 2ad = 2 * 9.8 * 2 = about 40 m2/s2 (for simplicity).
vf = sqrt(40) m/s = about 6 (for simplicity)

For an object that travels at 6 m/s, we can use equation (2) above to learn how much time it must decellerate over in order to experience 900 Gs. 900G is (approx) 9000 m/s2 so we get
t = vi / a = (6 / 9000)s = 0.00067 seconds

So what we learn is that a disk that is dropped from 6 feet has a speed of 6 m/s as it hits the ground. Upon hitting, it starts deforming and if this is a reasonably linear process (which may or may not be the case) then a constant 900Gs throughout means that it took 0.00067 seconds to come to a complete stop.

Determining how many millimeters of the disk and/or ground got deformed during the decelleration is left as an exercise for the reader :-)

So is this just a perf thing, or what?

[jonadab]

Other than potentially better performance stats, is this actually anything we need to know about?

To me it sounds like an implementation detail, which ought to be encapsulated behind the interface (SATA or whatever), so that nobody but hardware developers should ever actually need to know about it. Am I just weird, looking at it from a too-high-level perspective, or is this kind of detail completely irrelevant in practice?

Advantage Over RAM Cache?

[Paul Slocum]

After you boot up and things are cached in RAM, how does this help? And it seems like people reboot their computers so rarely these days anyway. How much time does it take to read 256MB from a disk into a RAM cache at bootup -- 10 seconds? Seems like your money would be better spent on more RAM, which is cheaper and can be written much faster.

Re:Advantage Over RAM Cache?

[MBCook]

This really isn't about reads but writes. By using this they can collect writes better (so they have to move the spindle less), cache the writes here (so they can avoid spinning up the disk longer), and protect writes (write in to this, power goes out, data still safe... RAM wouldn't do that). There isn't really much point to this for reads, as just sticking a little more cache (say 64MB) on the drive would work just about as well there.

Re:Advantage Over RAM Cache?

[tomhudson]

"and protect writes (write in to this, power goes out, data still safe... RAM wouldn't do that)."

flash memory doesn't do that, either. Ask anyone whose flash drive has died because they removed it prematurely. Abrupt power-down can kill flash dead. Heck, one guy at the office killed 2 flash drives in one day, simply because his install of XP is screwed up (so what else is new?) and even after the drives were supposedly safe to remove, they weren't. One I could believe as coincidence - 2 is a pattern.

Re:Advantage Over RAM Cache?

[MBCook]

Writing to flash is very fast. If the power fails during that, yeah, you're dead. But it will provide protection in between the periods of writing to flash and writing to disk. If things are bad enough (say, highly fragmented) this could be a decent amount of time. Basically, use the flash to help implement transactionality.

Re:Advantage Over RAM Cache?

[tomhudson]

Seems to me that journaling file systems already provide this, without the added risk of flash failure.

Re:Advantage Over RAM Cache?

[KaptajnKold]

For one thing, I don't believe there's actually any risk of flash failure, contrary to what you claim: I have a very limited knowledge of electronics, but I'm quite sure it's possible to implement a system using capacitors to saves a burst of power to perform whatever needs to be done, in order to avoid catastrophic flash failure.

But you're also missing the point entirely, if you think the flash cache is there to ensure data safety. As you are quite right in pointing out, that job is left to the journaling. The job of the flash cache as been pointed out repeatedly, is to increase preformance during write operations.

Continuous power?

[tepples]

if the OS knows that there are a few hundred [MB] of data to which it regularly needs access, why not cache them in regular system memory? Because every time the computer loses power, the cache will disappear. Unlike SDRAM, flash is nonvolatile. What's the price of this compared to the price of a good UPS?

Re:Continuous power?

[g0at]

Because every time the computer loses power, the cache will disappear. Unlike SDRAM, flash is nonvolatile. What's the price of this compared to the price of a good UPS? Hmm, I see what you're getting at. However, what would mediate which processes are worthy to write data to the NVRAM? I mean, no app wants to lose data.

b

Re:Continuous power?

[tcolberg]

On the contrary, based on how many times my Win98 install would crash per day, I would argue that it was eager to lose my data.

Re:if it's a hybrid

[EvanED]

You can't use DRAM for write caching; you can use flash for write caching.

The wearing out of flash is greatly overstated.

Re: Samsung not first to ship

[mckyj57]

Samsung only announced a prototype. Seagate is shipping, aren't they? I believe someone else is shipping as well, but the Seagate announcement is still significant.

Here is an alternate article for the slashdotted original:

CNN on Seagate announcement

The AC response isn't very clear

[Flying pig]

It isn't the six foot fall, it is the sudden stop. Just like going from 0 to 60 in a car in a mile or so is unnoticeable acceleration, but going from 60 to 0 in two feet, by hitting a brick wall and having the front crumple, is a bit noticeable (though for a short time only.)

Of course the deceleration depends on many factors, including the surface. In the days of metal case germanium transistors, it was well known that dropping a transistor three feet onto concrete could break an internal wire bond, while falling onto carpet would have no effect whatsoever. Interestingly, the 10lb object dropped on concrete will most likely experience less deceleration than the 1lb one - because it will deflect or crush the concrete to a greater extent. However, you are not advised to try this using a large SCSI drive and a 1.8 inch drive, because owing to the effects of scaling, the SCSI drive will be less rigid.

Re:The AC response isn't very clear

[QuickFox]

Just like going from 0 to 60 in a car in a mile or so For modern people using modern measurement units that's just below 100 km/h.

going from 60 to 0 in two feet, by hitting a brick wall Those two so-called "feet" are actually 60 cm.

Sheesh, I can't believe we're already well into the Internet age and still people keep insisting on using those antiquated units. "Feet". Why not "toes" and "noses"?

Re:The AC response isn't very clear

[Anonymous Coward]

Shut the fuck up.

What about that Cubic HD?

[HartDev]

I remember hearing something way back in the day about a hard drive that was to be a cubic inch and a terabyte of storage, something about a laser changing the composition or glass molecules.....can anyone help me out on this, or am I just making this up?

Re:if it's a hybrid

[doggiedoll]

Normally flash has a 10^7 erase/rewrite cycles. (from text book) It might be more for the new-age tech. But it is far lower than magnetic drive anyway.

hybrid?

[peektwice]

So does this mean it'll get better gas mileage in the city than it does on the highway?

hybrids

[Anonymous Coward]

There's been a few amphibious cars,which would classify as car/boat. The one with the coolest name is the volkswagen variant schwimmenwagen.

Benchmarker Mother Trucker

[Anonymous Coward]

Benchmarks, Where are they?! Performs this and that, menamena - bullshit. Who gives a smelly turd - show me graphs, tables and bars, not an Essay! Plzkthx.

Dual read/write heads?

[Theovon]

I've sometimes wondered... why don't drive manufactures make drives that have more than one read/write head arm? Random access times would be dramatically improved, no?

Cost? Reliability? With what they're spending to add flash, they could afford the cost of another head. Of course, it also occurs to me that if you want multiple heads, you can go for some kind of RAID configuration.

Re:Dual read/write heads?

[Anonymous Coward]

At first I thought you meant multiple heads per platter surface.

Anyways, does using RAID help random access times? I wouldn't think so. Your random access time would be the random access time of which ever drive is slowest. So it wouldn't affect it much, but maybe like 1ms or so.

Re:Dual read/write heads?

[Thundersnatch]

Anyways, does using RAID help random access times? I wouldn't think so. Your random access time would be the random access time of which ever drive is slowest. So it wouldn't affect it much, but maybe like 1ms or so.

That's true for a single random read access, but when you have a whole bunch of random accesses to do in a short period of time, the work is essentially divided amongst the drives in the RAID set, as the blocks are distributed randomly amongst the drives in the RAID set. Total throughput for random reads per second will essentialy be multiplied by the number of drives.

This read throughput inprovement holds in RAID 5 and RAID 0, and most other hybrid RAID modes such as RAID 6 and RAID 10. It is not generally true for write I/Os, which are more complex and time consuming depending on the RAID mode.

I think this is silly

[Anonymous Coward]

I'd rather have a 10GB 100% flash, then 256MB of several/hundred gigabytes.

Re: Samsung not first to ship

[Barnoid]

No, that's not correct. Samsung is shipping hybrid hard drives for over half a year now (see, for example, http://www.engadget.com/2007/03/07/samsungs-hybrid-hard-drive-hhd-released-to-oems/).

Recently, they even blamed Microsoft for the poor performance of hybrid hard disks on Windows Vista (in German, http://www.heise.de/newsticker/result.xhtml?url=/newsticker/meldung/97021&words=Samsung%20Hybrid&T=samsung%20hybrid)

Re:if it's a hybrid

[erroneus]

Theoretically, the most-read or otherwise booted-from files will be moved into the initial 256MB of space. This will enable that flash to boot the OS and most frequently used programs much faster. And since the amount of writing will be somewhat limited, the fact that flash tends to break down after 1 million or whatever writes makes it almost irrelevant.

But to address the notion that "DRAM" should be used, I'd have to say that is a silly notion. There are other types of RAM... RAM that doesn't need a constant refresh. For that, simple battery-backed RAM would be a good idea. And for that, I generally agree. I'd rather have other forms of RAM in there than flash.

But you know? If anyone wants to test some of this stuff out early, you can do this fairly inexpensively. Get an IDE to CF adapter device and a compact flash card. For 8GB pricewatch is showing about $90 and for a 16GB you're looking at around $140. For SATA users, you'd need yet another adapter... I've never heard of a SATA to CF adapter but I suppose it's conceivable. In any case, after that, you'd be able to see what it's like to run from flash over a mechanical drive. I imagine it would be impressive.

technology seems immature

[sentientbrendan]

The real world benefits of using flash as a cache layer between the harddrive and the computer, either through hybrid drives, don't seem to have materialized yet.

With my thinkpad there was an optional gig of flash that I ordered. After I downloaded the drivers and got it all set up, I found that there wasn't any noticiable difference in speed, or harddrive usage. However, I did notice that it interacted poorly with the "active protection" feature that stops the harddrive whenever the computer is in motion. Whenever the computer was unplugged, the flash cache was turned on, I could simply shake my computer (thus activating active protection) to get a blue screen.

Furthermore a little research showed that benchmarks on flash caches being sold right now offered no performance benefit whatsoever.

If there's no performance benefit, why are they trying to sell these things to people? I've seen some handwaving over the idea that flash *might* keep the harddrive from spinning most of the time and thus save battery life. However, when using the flash I saw no noticeable benefit.

Having an extra layer of cache in the system architecture seems like a good idea on paper, but in the real world the consumer is buying totally worthless pieces of hardware that do not improve performance one whit, and have never been proven to improve battery use.

Good to hear

[rk075245]

It is happy news, that every day the size of the hard disk getting smaller and smaller, So that we can avoid the huge space requirements for the data centers and also it will make it convenient for carry the lesser weight laptops.

gasoline??

[jmil]

LOL... some hybrid. sheesh.

And here I thought seagate just released a hard drive that can run on electricity OR gasoline.

Just *think* of the marketing: "Seagate's new 6 cylinder 4 platter 240 hp* 500 GB drive runs just 25 Gbpg (gigabits per gallon)** and goes 0->5400 in 3.85 seconds"

*1 horsepower = the counting power of 1 horse brain **1 gigabit = 1 billion bits.

only 256 meg

[ticktickboom]

perhaps they did the same with flash as they have done with platters, they always 'hid' a few secors, or whatever thier called. so when 1 goes bad theres a spare, so you never show bad sectors. the same with the flash. 256 is all an ide can read, or sata for that matter, the drives arnt that fast yet. when thier runnin at the 800mhz bus speeds, that would be nice. till then...they use tricks

Ahem, wrong cache type...

[fifirebel]

What you are describing is a write-back cache.
Brain fart? Or I misconstrued you?

Re:Ahem, wrong cache type...

[TheRaven64]

I'd consider it to be a write-though cache, since every write to the cache will correspond to a write to the disk, although, obviously that will take longer than the write to the cache. The only exceptions to this are in cases where the write delay is so large that the written data is already obsolete. Neither term is really designed for this kind of system, but I think it's closer to a write-through than a write-back model.

Permanent solid memory cache?

[garompeta]

Ok people, now my question is, how do we guarantee the effective wipe of our data from this hard drives?

Re: Samsung not first to ship

[andersa]

From TFA:

hybrid notebook disc drives that combine disc storage with flash memory to deliver ultra power efficiency, faster boot-ups and greater reliability for the exploding laptop PC market.

I am a bit surprised to find, that there is a market for exploding laptops..

An expensive RAM cache?

[noidentity]

The only difference between this and a RAM cache in the PC's memory is that this is preserved across power cycles. So the only benefit might be faster rebooting. Otherwise, why not just use some of your RAM as a hard drive cache? It can be more intelligent than the one on the drive since it has access to contextual information from the OS.

Re: Samsung not first to ship

[kestasjk]

It's a bit hush-hush at the moment, but basically the market was created when Bush announced the OLPT, One Laptop Per Terrorist, project.
Bush and his crazy hair-brained schemes.. I think it's tipped to be the next thing after the "surge" fails.

standard ATA but specific Windows ReadyDrive code

[spage]

The ideas behind this are applicable to any O.S. and there are proposed standard ATA commands to manipulate the Non-Volatile cache, see http://www.t13.org/Documents/UploadedDocuments/docs2007/D1699r4b-ATA8-ACS.pdf. I hope Linux and Mac hackers are working on it.

I'm not sure if the drive takes advantage of NV cache without specific O.S. support. Even without O.S. support, the drive could decide "You keep reading blocks X Y and Z, so I'll store them in NV cache" (drives already do this with their conventional RAM cache, typically 8MB) and "I'll keep your pending data writes in NV cache while waiting for the disk to spin up".

Windows Vista's ReadyDrive takes specific advantage of this feature: "During shutdown or hibernate all the disk sectors needed to boot or resume are pinned into the NV cache... Offsets within files and/or specific LBAs can be specified by the PC OEM in registry for pinning in the NV Cache". I converted the MS PowerPoint presentation http://docs.google.com/Present?docid=dcvvrqtp_13gj635t&fs=true (yay Google Docs, die PowerPoint DIE!).

sure there is

[r00t]

Dell and Sony sold a bunch of them.

OK, "venting with flame" if you want to get technical...

People buy them! Really!

Re:Does it make you smell your own farts and like

[Anonymous Coward]

Seagate missed an obvious opportunity in naming this thing. Instead of the "Momentus", they should've called it the "Pious", after a similarly-named Toyota hybrid consumer technology product. Then they too could've been behind the choking clouds of smug that have been forming above cities like San Francisco.

Re: Samsung not first to ship

[algoa456]

They are all the rage in the Middle East

So that's what "hybrid" meant

[haaz]

I was getting ready to post something about how a diesel-powered hard disk would be better in terms of long-term performance, the eventual need to replace batteries, and the ability to use alternative, non-gasoline fuels. But I realized then I realized, that's not what it's about.

Which is good, because as much as I like the smell of my car's biodiesel exhaust, I really wouldn't want my computer belching it out in my house all the time.

Re: Samsung not first to ship

[Anonymous Coward]

Or is that No Terrorist Left Behind?

Re: Samsung not first to ship

[GregNorc]

Real. Artists. SHIP.

Re: Samsung not first to ship

[GuyverDH]

http://www.samsung.com/us/consumer/subtype/subtype.do?group=computersperipherals&type=harddiskdrives&subtype=hybridhdd_flashon

That's the page with 3 hybrid drives listed...

"writing" vs. "erasing"

[Namlak]

By the way, flash has a slight weird characteristic that you can write to it with a byte granularity, but only erase it with a block granularity, and it's the number of erases that cause the problems.

Does "erasing" mean "make zero"? If so what's the difference between "erasing" at the block level and "writing" zeros at the byte-level? Is block-level writing/erasing faster than multiple individual byte-level writes? I'm sure it's trivial to implement some logic that can pick the more approriate operation.

Re:"writing" vs. "erasing"

[TheRaven64]

Erasing means you can write to it again. Initially, you can write to each byte in a block of flash. After each byte has been written, however, you can not modify the contents without erasing the whole cell. This loses the contents and puts the cell back in a state where you can write to it. Erasing is very different from setting to zero. After the erase cycle, the state may be zero, but it is a different kind of zero to before; it is one that you can set to one (actually, I seem to recall that the erased state for flash was one, but I could be wrong).

The first flash drive I owned was a Psion Solid State Disk, which cost £30 for 128KB, and was a single cell. You could create files on it until it was full (modifying files required rewriting them, and 'deleting' them just removed them from the FS view). After this, you had to 'format' the drive, which really meant erasing the flash cell.

Slashdot is US-centric

[Flying pig]

It's only polite to use the units of the country in which you are working, or for which you are writing.

I know it's a troll, but even so. I will make only two small points:

The SI system is as arbitrary as the Imperial system.

I use the SI system all the time, but it has one weakness. There is no convenient unit that results in small numbers for everyday use. It is convenient to say that, say, a table is 6ft by 3ft, the approximation involved being obvious. To say the same in SI, it has to be described as either 1800 by 900 mm or 1.8 by 0.9 metres. This involves either inconveniently large numbers that give a spurious precision, or the need to be familiar with decimals - which a lot of people are not. For scientific and engineering purposes, SI is the only sane way to go. For everyday use it is not ideal. This is because Imperial units are based on human scale, not geographic.

And, btw, your "cm" is not an approved SI unit. SI goes by thousands not tens.

Re:Slashdot is US-centric

[QuickFox]

Here in Sweden, for things like your table we use decimeters. A dm feels nicely human-scale because it's very roughly the width of a hand (with or without thumb depending on the hand). Your table would be 18 by 9 dm.

Strangely, lots of metric countries don't use the dm. I really can't see why. With mm, cm, dm and meters you've got very practical units for all sizes from small to big, and your choice of unit conveys a clear message about precision. You give a table's size in dm for rough size, in cm for greater precision, and in mm to convey the fact that you're giving the table's exact size.

And, btw, your "cm" is not an approved SI unit. SI goes by thousands not tens. In Sweden we use deciliters for recipes. Very practical, it's like a rather small coffee cup.

We even use a special kind of "mile" which is ten kilometers. Very handy for distances between cities.

From my Swedish standpoint I'd say the proper way to use metric units in everyday usage is to use what's practical. The SI is about science and technology. Let them stay in those fields. For everyday usage, use what's practical in everyday usage. But keep it firmly based on SI units, for convenient, easy conversion.

It's not even conversion. For me 18 dm and 180 cm mean the same size, only the precision is different.

I bet metric units would be more popular in the US and the UK if they were used with nicely practical units like dm and dl, rather than sticking too much to what the SI says, which is intended for science and should stay in science.

Ah, and by the way, I had no intention of trolling. It was just a joke with a glint in the eye. Occasionally after posting a joke I notice that my tone didn't come out quite as friendly as it was intended. A risky business, joking.