Alienware Puts 64GB Solid-State Drives In Desktops

Lucas123 writes "In the face of Seagate's announcement this week of a new hybrid drive, Dell subsidiary Alienware just upped the ante by doubling the capacity of its desktop solid-state disk drives to 64 GB. Dell has remained silent on the solid-state disk front since announcing a 32-GB solid-state option for its Latitude D420 and D629 ATG notebook computers earlier this year. Now, Alienware seems to be telling users to bypass hybrid drives altogether. 'Hybrid we consider to be a Band-Aid approach to solid state,' said Marc Diana, Alienware's product marketing manager 'Solid state pretty much puts hybrid in an obsolete class right now.'"

Comments

Re:what does it do to load times?

[evol262]

No idea who modded this 'underrated,' but those buses have nothing to do with this. The AGP bus never had any effect on storage performance (isolated), the PCIe bus is much faster than storage, etc. The IDE controller is on the Southbridge, and it's not bottlenecking. Storage is the bottleneck more often than not (seek times and raw speed). Will this cut down on seek times? Yes. Solid-state storage has nigh-instantaneous seek times, since there aren't any heads seeking.

Re:many write cycles?

[AKAImBatman]

What if the OS decides to write stuff to certain sectors all the time?

Most flash controllers remap the sectors on the fly to ensure that the memory is not worn down prematurely. So if you rewrite the same logical sector 5 times over, a chance exists that you'll get 5 different physical sectors.

Re:many write cycles?

[tygerstripes]

This is an old /. topic, really. Key points:

1. Flash used to have a limit of about 500,000 read/writes. That limit has since been surpassed. I gather it can exceed 1 million now, though Wikipedia still says the former.
2. Although it wasn't addressed in the article (dammit), it has often been suggested that some on-disk monitoring and allocation mechanism will prevent areas from burning-out, or from being used if they do burn out. (This will be a particular issue for page/swap/scratch-files)
3. Given that hard drives usually have a MTBF of something like 3-5 years, the technology only has to be good enough to meet that standard before it becomes as technically viable as HDDs.
4. Given its other advantages over existing HDDs (even hybrids), I imagine that it will be considered viable - especially in laptops - long before it reaches that level of robustness.

Can I just say, it's about time they brought out a version that could compare with existing low-end laptop drives in terms of capacity. If you ask me, that's what was really holding back the big-spenders from buying into this tech.

Re:what does it do to load times?

[skulgnome]

The PCI bus isn't a bottleneck until you start getting over 120 megs a second down from a hard disk. Basic parallel PCI transfers up to 133 megs per second, theoretical, and even a single lane of PCI-e is quicker than that.

Re:Eventually

[morgan_greywolf]

Yeah, but as the first adopters and the die hard gamers looking for every advantage they can get buy more of these, we'll see the price drop eventually. Sorry to burst your bubble, but eventually >= 10 years. Flash memory is expensive to produce, and production of flash memory already outstrips HDDs (think of all those USB thumb drives).

Re:many write cycles?

[AKAImBatman]

The term is "Wear Leveling", and it's built into standards like SD Cards. Doing a quick Google search produces white papers like this one:

http://www.stec-inc.com/downloads/AN-0702_STEC_SMALL_CARDS_WEAR_LEVELING_LIFETIME_CALCULATOR.pdf

Re:many write cycles?

[glwtta]

Given that hard drives usually have a MTBF of something like 3-5 years

Pet peeve: MTBF is not life expectancy, it's the average time between failures if you replace the drives before they are expected to die. Common MTBF are currently anywhere between 50 and 150 years (mostly made up numbers), whereas life expectancy is in the 3-5 years range (at best).

Great on Battery life

[Poppageorgio]

I have a Latitude D430 for work with a 32GB SSD, and while it isn't noticeably faster than the guy next to me that has a standard HDD in the same machine, my battery life is WAY better. I'm getting 10+ hours with the extended battery out of the thing. And, I'm not as scared about losing data due to a dropped laptop. (Networking = frequently dropped laptops!)

Re:Great on Battery life

[btSeaPig]

Well, I have a Latitude D630, with the 32GB SSD, running Ubuntu 7.10. OpenOffice opens in right at one second. Very impressive if you ask me.

Re:Have they solved the longevity issue?

[plague3106]

You have a flaw in your theory though; a portion of the drive won't be changing much, because OS and program files don't change too much. So there's a part of the disk that is only written to rarely, and other parts of the disk that will be written to more often, because a chunk of the drive won't change.

So while the drive would still last a long, long time, you do need to keep in mind the above.

Re:what does it do to load times?

[networkBoy]

Single lane PCIe is 1.25Gbps.
After you move to bytes and remove overhead you get 150 MBps.

Can we let the old "write limits" go now?

[Overzeetop]

I don't get it. Modern flash has 1M+ write cycles, and we might presume that there is some rudimentary write balancing in these drives. If you work 1GB of cache (not unlikely, and probably on the low side for Vista), I get 8Gb x 1M writes = 8x10^15 write operations before your 1GB area fails completely. Using load balancing, and dynamic reallocation of a 64GB disc, but taking the "limit" of useability at 50% of the write cycles before you are might start to worry, how long does it take to write 256x10^15 bits (8x10^15 x 64GB x 50%)? Well, TFA didn't give write speeds, so I'm going to presume a ludicrous write speed of 50MB/s (I'm not aware of any consumer-grade flash that writes that fast). 50x8=400Mb/s or 4x10^8 b/s. So if I've got my exponents correct, that put the 50% threshold at an even 64x10^7 seconds, or about 177,777 hours of continuous writes, or only about 20 years. That presumes you actually have your machine (a) never reading the cache, and (b) never writing anything else to the disk, since the entire bandwidth taken up by the cache writing and (c) it's doing this 24/7 (as I presume Vista attempts to do).

And at this point, your drive will be through 50% of it's theoretical write-cycle life. And about 1/1000 the capacity of the drive you would be able to buy for $100 to replace it.

Re:Can we let the old "write limits" go now?

[pslam]

I don't get it. Modern flash has 1M+ write cycles, and we might presume that there is some rudimentary write balancing in these drives.

Strangely enough, modern flash is about 100k write cycles for high density SLC NAND and 10k writes for MLC NAND. Newer flash actually gets worse as the densities get better.

Even so, with proper wear leveling and sufficient redundancy you can achieve failure rates better than a spinning media. In fact, you can pick the numbers to achieve any arbitrary failure rate.

As for speed - you're correct, no single flash chip is 50MB/sec, but you can stack many of them in parallel and get that. That's a common way of doing it.

I think you're being overly harsh and pessimistic with your figures. There are some workloads you obviously shouldn't pair with a NAND flash, but quite frankly gaming isn't going to stress these things.

Re:Can we let the old "write limits" go now?

[Overzeetop]

Well, I googled and came up with this: http://www.baydel.com/images/gallery/NAND%20flash%20resilience.pdf

which is a whitepaper which shows a 64GB NAND device with a 100MB/s write speed can go for 20 years continuously, with error correction, before hitting the write limit. They didn't use the same numbers, but for a device with a likely lifetime of less than 5-6 years, they certainly seem to be up to practically any task.

Re:what does it do to load times?

[pslam]

What nobody is pointing out is that a standard windows install will thrash the hell out of a Solid state drive. There is a reason you need to balance your writes and not treat a SS disk like a hard drive. I destroyed a Solid state IDE drive back 6 years ago (you have been able to buy them for over 15 years now) by installing windows on it. the swap space died within weeks.

These days (well, since YEARS ago now) we have this thing called Wear Leveling which means you can't wear out NAND flash by simply writing over the same portion over and over again. The writes get spread around other areas instead.

It hasn't been possible to kill a (decent) solid state drive like this in a very long time now. Please don't misinform people.

Re:obsolete?

[morgan_greywolf]

You mean *per gigabyte* and that's true. But tape drives are even cheaper No, they aren't. Go look on Pricewatch for an LTO Ultrium drive. LTO4 drives are about $1100, with the older, lower capacity LTO2s being about half that. The tapes themselves might be lower than HDDs in terms price/GB, but then they don't last nearly as long as an HDD, either.

Anyway, the biggest problem with tapes is that they aren't a random-access media. That's why they aren't used as a means of primary storage.

Re:Dell offers 128GB on XPS M1730 notebook

[JPEWdev]

The 128 GB Drive in the XPS M1730 is a 2x64 GB Raid, so it is not really more advanced technology.

Re:many write cycles?

[AKAImBatman]

WTF are you talking about? Wear leveling is usually paired with ECC to prevent exactly the types of issues you're talking about. The technology is almost exactly the same as ECC hard drive technology. Detect a bad sector, mark it as bad, remap the sector, rewrite data to new sector. Rinse and repeat.

Naive storage devices like you describe haven't been common for quite a few years now.

Re:Better Data Security?

[aegl]

Does flash technology leave a phantom image after it's erased like magnetic storage does?

Almost certainly yes. Lots of people have been asking about the number of write(erase) cycles that a flash drive can perform, and many others have been answering "wear levelling". This means that when you re-write block 100000, the flash controller will actually write the data to some other physical block and just internally renumber that new block as 100000. This of course means that the data you "overwrote" is still in the original location.

Because you have no control over the wear levelling algorithm, even writing to the whole drive may not remove all traces of old data. The drive may internally have a larger than rated capacity with plenty of spare blocks to use as replacements for ones that reach their wear level limit.

-Tony