Slashdot Mirror
"Limited Edition" SSD Has Fastest Storage Speed
Vigile writes "The idea of having a 'Limited Edition' solid state drive might seem counter-intuitive, but regardless of the naming, the new OCZ Vertex LE is based on the new Sandforce SSD controller that promises significant increases in performance, along with improved ability to detect and correct errors in the data stored in flash. While the initial Sandforce drive was called the 'Vertex 2 Pro' and included a super-capacitor for data integrity, the Vertex LE drops that feature to improve cost efficiency. In PC Perspectives's performance tests, the drive was able to best the Intel X25-M line in file creation and copying duties, had minimal fragmentation or slow-down effects, and was very competitive in IOs per second as well. It seems that current SSD manufacturers are all targeting Intel and the new Sandforce controller is likely the first to be up to the challenge."
Is the cap left off the board so you can just put one in yourself or is it size-reduced as well?
"I was so eager to test it that I pounded on this drive all night "
Possible poor choice of words?
Orbis terrarum est non altus satis
Is that new-speak for "cheaper"? I also love "the drive was able to best the Intel X25-M" this is one of the worst written pieces of commercial press release I have ever seen on Slashdot.
"to improve cost efficiency"
should be
"to lower the cost"
Because we're talking about the home/enthusiast market, which is completely different (including and especially in price point) from the enterprise storage market.
Why does it matter if they get their blazing fast speed by fragmenting all the data all over the place? On hard disks fragmentation is a bad thing, on SSDs it's a good thing, what's your point?
It's just like LCDs getting their amazing thinness by having individual pixels instead of scanning an electron beam across a vacuum tube.
Disgusting!
If "almost a halt" is 200MB/s read speeds as opposed to 260, I think I can live with it before I upgrade to my TRIM firmware, which negates the whole issue... whoops, I started using TRIM on my home drives months ago.
Seriously, the SSD market has exploded in the last 12 months. It's gone from being an expensive tool useful to enthusiasts to a not-quite-as-expensive-but-faster-than-any-number-of-hard-drives-can-provide utility that's worth five times it's price, especially for enterprise users.
* Proud owner of 1 intel SSD, 3 OCZ SSD's and administrator of about 3TB of SSD SAN and >8GB FusionIO cache with a bunch of spinning magnetic domains in the background that we can't get rid of fast enough
Moderation Total: -1 Troll, +3 Goat
I've got a question here, if you don't mind me asking:
Are SSDs more prone to errors than disk drives? If so, why?
There seems to be some strangeness about SSDs and if I try to go read some technical papers on them on a Friday night when I've half a snoot full, it's going to make me all headache-y.
And regarding this "trim" function, can't they just make the nodes smaller?
And is this problem just going to go away once they get the manufacturing capacity to make gigantic SSDs the way they make gigantic hard drives? I remember when there was a lot of effort to be really efficient in the use of hard disks and now it doesn't seem as important when you've got 2TB drives on sale.
You are welcome on my lawn.
Calling the article a 'press release' unfairly tarnishes OCZ. Their press release is still full of press release though:
http://www.ocztechnology.com/aboutocz/press/2010/362
Nerd rage is the funniest rage.
Why just for "enterprise" users, and what does Star Trek have to do with it anyway?
Would SSDs make a big difference for people who create and edit sound or video? If you tell me it'll improve the performance of my digital audio workstation or video editing software, I'll blow my fat tax refund check on some SSDs right now. Can I just hook up SSDs to the SATA controller on my machine?
Hell, I'm just full of questions. Irish coffees tend to do that to me.
You are welcome on my lawn.
I've had one in my laptop for about 8 months and write gigabytes to it every day, particularly suspending VMWare images to disk. It still writes at 140 MB/s sustained (to ext3 filesystem, not just raw write speed). That might be slower than when it was new, I don't remember, but it destroys any laptop harddrive. This drive was expensive though, like $800 IIRC, but it also supports full-disk hardware encryption which was mandated at my workplace.
Before that I had a first-gen X25M. It did slow down more, but still completely blew away hard drives. "Slowing down to almost a halt," no, not even close. Especially for multitasking, which brings HDDs almost to a halt.
As you can see for this newer drive, there is practically no slowdown, and in any case even its slowest results are many times faster than any laptop HDD.
I've heard they're working with a Trim function thingy to remedy this, but I haven't really paid attention since.
If you're going to take the time out to post and bitch, at least read up and know what you're bitching about. They've had TRIM for a while now, and Indilinx firmware can collect fragmented nodes during pause time.
These are solved problems you're bitching about.
What do you do for an encore? Complain about how annoying it is to get across town in a horse and buggy?
Weaselmancer
rediculous.
Up until 2003, I used an old PowerMac as my main machine, running Mac OS 7.6.1. I kept the system folder on a RAM disk, and booted off that. Blazing fast, but had the bad habit of losing all data whenever there was a power failure (nothing a periodic mirror-to-disk couldn't remedy, though). I still use that old PowerMac all the time (running Mac OS 9.2.2 now) and keep a persistent 130 MB RAM disk for storing temporary files. It persists across reboots, though I don't keep the system on it anymore.
The new OCZ SSDs, while a welcome addition to the market aren't anywhere near "fastest storage".
Crucial RealSSD C300: http://www.tweaktown.com/reviews/3118/crucial_realssd_c300_256gb_sata_6gbps_solid_state_disk/index5.html
Fusion-IO: http://storage-news.com/2009/10/29/hothardware-shows-first-benchmarks-for-fusion-io-drive/
If you're doing non linear sound and video editing with multiple simultaneous streams coming out of many files, and those files are greater than the amount of your available RAM (common), then yes, an SSD would make a big difference. You'd also experience comparatively blazing boots on your workstation. And yes, SSDs will connect to your SATA controller. As far as the system is concerned, they are hard drives. Very very fast hard drives.
I'm kinda fed up waiting for the SSD manufacturers to get their act together. There's just no reason for drives to be only 10-50x faster than physical drives. It should be trivial to make them many thousands of times faster.
I suspect that most drives we're seeing are too full of compromises to unlock the real potential of flash storage. Manufacturers are sticking to 'safe' markets and form factors. For example, they all seem to target the 2.5" laptop drive market, so all the SSD controllers I've seen so far are all very low power (~1W), which seriously limits their performance. Also, very few drives use PCI-e natively as a bus, most consumer PCI-e SSDs are actually four SATA SSDs attached to a generic SATA RAID card, which is just... sad. It's also telling that it's a factor of two cheaper to just go and buy four SSDs and RAID them using an off-the-shelf RAID controller! (*)
Meanwhile, FusionIO makes PCI-e cards that can do 100-200K IOPS at speeds of about 1GB/sec! Sure, they're expensive, but 90% of that is because they're a very small volume product targeted at the 'enterprise' market, which automatically inflates the price by a '0' or two. Take a look at a photo of one of their cards. The controller chip has a heat sink, because it's designed for performance, not power efficiency!
This reminiscent of the early days of the 3D accelerator market. On one side, there was the high-performing 'enterprise' series of products from Silicon Graphics, at an insane price, and at the low-end of the market there were companies making half-assed cards that actually decelerated graphics performance. Then NVIDIA happened, and now Silicon Graphics is a has been because they didn't understand that consumers want performance at a sane price point. Today, we still have SSDs that are slower that mechanical drives at some tasks, which just boggles the mind, and on the other hand we have FusionIO, a company with technically great products that decided to try to target the consumer market by releasing a tiny 80GB drive for a jaw-dropping $1500. I mean.. seriously... what?
Back when I was a young kid first entering university, SGI came to do a sales pitch, targeted at people doing engineering or whatever. They were trying to market their "low-end" workstations with special discount "educational" pricing. At the time, I had a first-generation 3Dfx accelerator in one of the first Athlons, which cost me about $1500 total and could run circles around the SGI machine. Nonetheless, I was curious about the old-school SGI machine, so I asked for a price quote. The sales guy mumbled a lot about how it's "totally worth it", and "actually very cost effective". It took me about five minutes to extract a number. The base model, empty, with no RAM, drive, or 3D accelerator was $40K. The SSD market is exactly at the same point. I'm just waiting for a new ''NVIDIA" or "ATI" to come along, crush the competition with vastly superior products with no stupid compromises, and steal all the engineers from FusionIO and then buy the company for their IP for a bag of beans a couple of years later.
*) This really is stupid: 256GB OCZ Z-Drive p84 PCI-Express is $2420, but I can get four of these 60GB OCZ Vertex SATA at $308 each for a total of $1232, or about half. Most motherboards have 4 built-in ports with RAID capability, so I don't even need a dedicated controller!
I have no problem with OCZ releasing press releases, they're a company that sells stuff so that's what they do. Slashdot OTOH is supposed to be some sort of quasi-news site (or at least it used to be) with discussion, not a PR mouthpiece.
I think Seinfeld, rather, had it right. "Limited to what, how many you can sell?"
Make sure you get either an Intel X25-M (though the biggest one they offer is 160GB) or something with an Indilinx controller (OCZ Vertex, for example, up to 256GB). Stay away from anything with a JMicron controller - those drives might be cheaper for bigger sizes, but the performance is crap.
Right, but this isn't PR, PC Perspective thinks they are a news site (and they didn't simply parrot the OCZ press release).
Nerd rage is the funniest rage.
SSDs are improving very quickly. Yes they are more prone but that isn't the main story. We know much more about regular HDDs and their failure modes than we do for SSDs. Because of this SSDs seem to fail without warning, or don't degrade as nicely. Which is more of the problem tbh.
Irish coffee's bring out the best in everyone ;)
Reason I started using them at home was due to video editing - not very useful for encoding when you can rarely outpace your CPU's capability to encode stuff, but for random seeking/non-linear stuff/extracting streams/muxing, SSD's are a boon. Depending on your workload you can even get away with using crappy SSD's that are shit at random workloads but awesome at sequential.
TBH though you'll get the most noticeable improvement with using it as your system drive; apps start almost instantly and there's never any thrashing as $bloaty_app loads. Heck, my linux machines boot in 5s with the comparatively cheap OCZ Agility drives; the difference is less noticeable in windows however. Try running a laptop off an SSD for a month and then go back to a mechanical drive - the apparent slowness will drive you crazy :)
The benefits for enterprise users are especially large because 20k of SSD can replace 100k of fibre channel whilst getting 10x the performance and greater reliability. Plus Picard totally loves SSD's as he can rest his tea, earl grey, hot, on them without risking Data loss.
Moderation Total: -1 Troll, +3 Goat
You can hook it up to the SATA.
Though if you have a spare PCIe slot, that would probably give you more throughput. Of course, the model that plugs into PCIe is different from the SATA model.
It's likely you would notice the benefit regardless.
This is computer stuff, so "Limited Edition" is more likely to mean: "After a few months when we need something 'new' for marketing reasons, we'll just add the super capacitor, call it the 'Pro' edition, and phase out the 'Limited Edition'".
At least not the Colossus I bought. Write speeds are great but read speeds suck compared to the Intels. The Colossus doesn't even have NCQ for some reason! There's just one tag. The Intels beat the hell out of it on reading because of that. Sure, the 40G Intel's write speed isn't too hot but once you get to 80G and beyond it's just fine.
The problem with write speeds for MLC flash based drives is, well, its a bit oxymoronic. With the limited durability you don't want to be writing at high sustained bandwidths anyway. The SLC stuff is more suited to it though of course we're talking at least 2x the price per gigabyte for SLC.
--
We've just started using SSDs in DragonFly-land to cache filesystem data and meta-data, and to back tmpfs. It's interesting how much of an effect the SSD has. It only takes 6GB of SSD storage for every 14 million or so inodes to essentially cache ALL the meta-data in a filesystem, so even on 32-bit kernels with its 32-64G swap configuration limit the SSD effectively removes all overhead from find, ls, rdist, cvsup, git, and other directory traversals (64-bit kernels can do 512G-1TB or so of SSD swap). So its in the bag for meta-data caching.
Data-caching is a bit more difficult to quantify but certainly any data set which actually fits in the SSD can push your web server to 100MB/s out the network with a single SSD (A single 40G Intel SSD can do 170-200MB/sec reading after all). So a GigE interface basically can be saturated. For the purposes of serving data out a network the SSD data-cache is almost like an extension of memory and allows considerably cheaper hardware to be used... no need for lots of spindles or big motherboards sporting 16-64G of ram. The difficulty, of course, is when the active data-set doesn't fit into the SSD.
Even using it as general swap space for a workstation has visible benefits when it comes to juggling applications and medium-sized data sets (like e.g. videos or lots of pictures in RAW format), not to mention program text and data that would normally be throw away overnight or by other large programs.
Another interesting outcome of using the SSD as a cache instead of loading an actual filesystem on it is that it seems to be fairly unstressed when it comes to fragmentation. The kernel pages data out in 64K-256K chunks and multiple chunks are often linear, so the SSD doesn't have to do much write combining at all.
In most of these use-cases read bandwidth is the overriding factor. Write bandwidth is not.
-Matt
This is completely backwards. It is hard drives which fail without warning. See Google's recent paper on the futility of S.M.A.R.T. And when an HDD fails, your data is _gone_. The best you can hope for is spending huge amounts of money to put the platters into another drive and reading the data back. The predominant failure mode for flash is erase cycle endurance exhaustion, upon which time the flash reverts to being read-only. Compared to a HDD the flash failure mode is hugely desirable. You can also monitor an SSD and replace it when it reaches the 100,000 erase cycle limit (or 10,000 for MLC). HDD has no such mechanism.
I don't see why manufacturers prefer spending large amounts of time and money into producing smart controllers when they could just give us raw access to the device and let us use something as NILFS on top of it. Do you?
Anyone else agree that SSD speeds are plenty fast for the tasks given to it? When I shop for SSD drives I look for a reputable company that doesn't stutter like crazy with reads and writes for the lowest price. I've owned Intel X25Ms as well as other brands and I can't tell the difference in performance. Of course, the benchmarks do paint different numbers.
But who is REALLY gonna notice that 0.03ms difference in "seek time" for one SSD over another and 150MB/sec over 220MB/sec sequential? SSDs these days are so fast I don't see a reason to "upgrade" to a faster SSD if I already have one.
What do I want to see improved on SSD? Reliability and price. This "Limited Edition" seems like a waste, and I'd bet that less than 1% of users here at slashdot would really and truely notice this. I'd bet most of us would be unable to tell the difference if tested blindly.
I'm sure this will hurt my karma, but I can't believe that I'm alone in thinking this.
Maybe in 1995.
I think I would prefer MILFS on top, don't you?
>8GB FusionIO cache with a bunch of spinning magnetic domains in the background that we can't get rid of fast enough
Is that supposed to be TB? Don't ioDrives come in 160GB multiples?
Mind you, if I had 8TB of ioDrives, there'd be no need for anything else. Each one of those has read speeds of close to 1GB/sec, and enough IOPS to beat a dozen of the next best competitor. Now if only they cost 15x less per GB.
Try running a laptop off an SSD for a month and then go back to a mechanical drive - the apparent slowness will drive you crazy :)
Not to mention the battery life decrease, HD -> SSD got me 40% longer battery life on my netbooks. About 11 hours in total now, which is the way it should be. Plus no more worries about vibrations, decreased heat and it's quieter.
Except that most modern SSDs actually have a 50,000,000 erase cycle limit, not 100,000. For reference, an X25-M writing continuously as fast as it could, constantly, wouldn't hit this until 140 years.
The other nice thing here is that with hard disks, the risk disk failure is constant when compared to the capacity of the device, with SSDs, the risk halves as the capacity doubles, because the controller can spread its writes out more.
SSDs really are *massively* more reliable than HDDs. They last for longer (assuming our estimates of their tollerances are right), their failure modes are good, and their chance of losing lots of data goes down the more data you squeeze in there. The only thing HDDs have in their favour is that we know roughly that they tend to fail in the 4-8 year window, so we should probably replace them at 3 years or so.
Personally, I'd be using something like YAFFS (Yet Another Flash File System) rather than NILFS.
Forget thrust, drag, lift and weight. Airplanes fly because of money.
It should be illegal to label products like this. The only thing limited, is the mental capacity of those who buy it because of this label. ;)
Any sufficiently advanced intelligence is indistinguishable from stupidity.
Thanks, MrNemesis. That's exactly what I'm going to do. I'm happy with the data throughput that SATA drives with big caches give me for streaming samples or video, but it would be great to have my system a little peppier.
I'm going to haunt the online stores later today, as soon as I get some breakfast.
You are welcome on my lawn.
Not exactly. In this kind of market it usually means 'we've got a better technology that we were hoping to get to market by now, but we had some delays. In the meantime, we worked out how to make the older tech a bit faster, so you can buy that while we fix the bugs keeping the new tech from the market. When we've finally got the new tech working, this line will look really overpriced because it's much more expensive than the newer design to produce and doesn't provide any significant benefits, so we'll discontinue it'.
I am TheRaven on Soylent News
This is completely backwards. It is hard drives which fail without warning.
I hate to break it to you but SSDs also fail without warning and go completely dead. Visit the PC Enthusiast forums and do some searches. 100% dead SSDs are very common despite their short time in the market. It's scary how many of these things are going belly up after only months of use.
Platter based disks and SSDs both have big complex controller boards and those controllers are subject to failure. The only thing you're missing with SSDs is the risk of mechanical failure. If a platter based disk is treated properly through it's lifetime there is little risk of mechanical failure until it is quite old.
I was going to grab a few 30GB SSDs when they dropped south of $100 but the failure rate has caused me to put my plans on hold. I'm going with 500GB per platter mechanicals for the time being. RAID a couple and you have sequential rates in the ballpark of the affordable SSDs.
"Liechtenstein is the world's largest producer of sausage casings, potassium storage units, and false teeth."
I see a few reasons
1: Most people will be running either windows which pretty much means you use NTFS whether you like it or not (or you could use fat32 but that isn't exactly going to be any better). Even on linux i'd imagine the number of clueless newbies who would set up a standard filesystem on the device and quickly ruin it would be pretty high. This means high RMA expenses and pissed off users.
2: putting the wear leveling control on the drive puts the drive manufacturer in control of it. That means they can tweak it to match the particular memory array they have, implement new developments without waiting for the operating system to catch up and so on.
3: exposing the memory array directly would require a different paradigm on the interface than current drives use to expose things like the difference between write blocks and erase blocks. Without a major change in interface specs this would mean the filesystem would have no way to get the true characteristics of the array.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
The predominant failure mode for flash is erase cycle endurance exhaustion, upon which time the flash reverts to being read-only. Compared to a HDD the flash failure mode is hugely desirable.
At my company a few years ago we purchased two OCZ SSD drives (using the now infamous JMicron controllers). They were for two identical systems, but we kept having problems with the first one we were setting up (using Linux). Everything would seem fine at first, but the system would start crashing and become unbootable within hours. We formatted and started over a couple of times replacing various pieces of hardware. Eventually we narrowed it down to the SSD by using a command (I forget the command used in Linux) to fill the drive with specified patterns of bits and then read back the data to see if it is correct. There was a patch where every other bit would not flip from 0 to 1. Of course, due to wear leveling the location of the patch would move around for each test, but it was easily reproducible in size. The drive was silently failing on write, producing random garbage. OCZ ended up replacing the drive, but we ended up not using them due to other performance issues. (OCZ has hidden the relevant threads on their forums since then.)
Anyway, that's my anecdote for the day.
So in other words you bought some cheap crap from a fly-by-night called "OCZ" and installed it in your servers, and were surprised when it didn't work. Do you also buy hard drives from no-name Taiwanese, or do you stick to Seagate, Western Digital, and Hitachi?
Crap is crap, whether HDD or SSD. And for that matter, memory. You can buy some janky erroneous DRAM from OCZ as well, but I don't recommend it.
I'm using the Patriot Torqx m28 that I got at Fry's. Peppy doesn't begin to describe it. I'm seeing 8800 small random read IOPs with Iometer, and 28000 sequential. Compare this with about 180 IOPs for a 15k SAS hard drive - it's over 40 times as fast. Boot time is well under 30 seconds. My Core2duo laptop is usable for work again, and I can finally work with virtual machines in a reasonable manner.
Any one you get is going to be better than spinning disk, but the newer ones really are much better and more durable.
Help stamp out iliturcy.
Photoshop 1.0 actually ran on a B&W Mac? Seriously? What's the point in that?
Although, if anyone know where I can find a copy of this for my Mac Plus, let me know...
So in other words you bought some cheap crap from a fly-by-night called "OCZ" and installed it in your servers, and were surprised when it didn't work.
The same "cheap fly-by-night" that this review is about?
Maybe you should pay some attention ;) The current generation of SSDs are all supporting some type of trimming.
I would expect the first generation of any technology to have issues. Don't write them off just because the first round was sub-optimal!
Get a descent standalone crypto accelarator, if your that short on CPU, and use a plain vanilla hard drive/SSD.
I know tobacco is bad for you, so I smoke weed with crack.