Slashdot Log In
Intel's First SSD Blows Doors Off Competition
Posted by
ScuttleMonkey
on Mon Sep 08, 2008 01:49 PM
from the bigger-smaller-deal dept.
from the bigger-smaller-deal dept.
theraindog writes "Intel is entering the storage market with an ambitious X25-M solid-state drive capable of 250MB/s sustained reads and 70MB/s writes. The drive is so fast that it employs Native Command Queuing (originally designed to hide mechanical hard drive latency) to compensate for latency the SSD encounters in host systems. But how fast is the drive in the real world? The Tech Report has an in-depth review comparing the X25-M's performance and power consumption with that of the fastest desktop, mobile, and solid-state drives on the market."
Related Stories
[+]
Intel Takes SATA Performance Crown With X25-E SSD 164 comments
theraindog writes "We've already seen Intel's first X25-M solid-state drive blow the doors off the competition, and now there's a new X25-E Extreme model that's even faster. This latest drive reads at 250MB/s, writes at 170MB/s, and offers ten times the lifespan of its predecessor, all while retaining Intel's wicked-fast storage controller and crafty Native Command Queuing support. The Extreme isn't cheap, of course, but The Tech Report's in-depth review of the drive suggests that if you consider its cost in terms of performance, the X25-E actually represents good value for demanding multi-user environments."
[+]
Will 2009 Be the Turning Point For SSDs? 290 comments
Iddo Genuth writes "Since first entering the consumer market about two years ago, solid state drives (SSDs) have improved significantly. While prices remain substantially higher than conventional magnetic storage, it is predicted that in 2009 SSDs will finally make an impact on both the consumer and business markets bringing blazing fast speeds at reasonable prices for the first time — will it finally happen?"
It seems likely, as Samsung began mass-producing both 128GB and 256GB SSDs this year. Intel and Micron have also posted recent breakthroughs which will help to bring the technology into the mainstream.
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
Oh Yeah? (Score:5, Funny)
My SBDs will blow THEIR doors off.
Re:Oh Yeah? (Score:4, Funny)
That's the beauty of it. You will never know!
Parent
Re: (Score:3, Informative)
--sabre86
Re:Oh Yeah? (Score:5, Funny)
That's not pedantry, it's rigour.
Thank you for taking the time to correct my misuse of a sixty year old acronym in an off-hand quip replying to a fart joke. It is this level of attention to detail which makes slashdot what it is.
Parent
Well, a step in the right direction (Score:5, Insightful)
A step in the right direction, but at $600 per 1000 I am gonna wait a bit longer before jumping on the SSD bandwagon.
Re:Well, a step in the right direction (Score:5, Funny)
Why? They're almost free at 60 cents each :-P
Parent
Re:Well, a step in the right direction (Score:5, Interesting)
A step in the right direction, but at $600 per 1000 I am gonna wait a bit longer before jumping on the SSD bandwagon.
I'd place an order for one this instant if I could. My company uses a relatively small database, on the order of 40GB of online data. It's running on 4 SCSI-320 Cheetah 32GB, 15K RPM drives in RAID 0. By all accounts, this single SSD would out-seek the Cheetahs, meaning that our website can serve more customers and more quickly. This is a total no-brainer for a lot of applications, even at the current price.
Parent
Re:Well, a step in the right direction (Score:5, Informative)
Before rushing to buy these for database use, I would want a good look at MTBF values. Especially MTBF values for really heavy use, which may be completely different from estimated desktop use.
Parent
Are you sure? (Score:3, Insightful)
Quote
4 SCSI-320 Cheetah 32GB, 15K RPM drives in RAID 0.
End Quote
What company would really want to run their DB on a Raid 0 (Striped) Disk setup? Does this not put it at risk from a single spindle failure?
Re: (Score:3, Insightful)
What company would really want to run their DB on a Raid 0 (Striped) Disk setup?
One who replicates the data to slower backup systems.
Does this not put it at risk from a single spindle failure?
If those were the only spindles involved, sure.
Re:Well, a step in the right direction (Score:4, Informative)
It's running on 4 SCSI-320 Cheetah 32GB, 15K RPM drives in RAID 0.
I hope you know how volatile RAID 0 can be. A problem with any single one of those drives will screw up the whole works until you can restore from a backup. I can understand wanting to avoid RAID 5/6 if there are a lot of writes to your DB as performance of those arrays in writes are notoriously bad and RAID 1 would be a doubled hardware cost increase, but the ability to stay up and hot swap in drives after a failure is priceless.
Parent
Re: (Score:3, Interesting)
I hope you know how volatile RAID 0 can be.
Oh yeah, but we can do a bare-metal recovery in an acceptable amount of time, so a failure is more along the lines of "dangit, break out the tapes".
To answer other posters while I'm at it:
That chassis is maxed out on RAM. We could buy a newer, bigger system but this SSD would serve about the same ends for a lot less money and effort. Besides, at some point you have to flush those cached writes out to disk. Right now, that is sometimes a bottleneck on our system. If we could magically make those writes s
Re: (Score:3, Interesting)
Re: (Score:3, Interesting)
Re:Well, a step in the right direction (Score:4, Interesting)
I hope you know how volatile RAID 0 can be. A problem with any single one of those drives will screw up the whole works until you can restore from a backup
Oh my, pardon me, I am rolling on the floor laughing, biting the carpet and frightening the cat (ROFLBTCAFTC).
I remember reading these exact same arguments in articles written during the early days of computing, when people were complaining of the multi-platter nature of modern disk packs. These started hitting the market around 1963 I think. The argument went -- if you stack all those platters together, the failure of one platter would trash the entire set! Oh noes...
Parent
Re:Well, a step in the right direction (Score:4, Insightful)
And...what? It doesn't?
Parent
Re:Well, a step in the right direction (Score:5, Insightful)
Have you tried just putting 16GB of RAM in the database server? Nearly 16GB of cache for a 40GB database should work pretty well.
More geenrally, it's time to start thinking about DB servers that satisfy all reads from memory. It won't be long before the RAM available in a commodity sever is larger than many shops' database. Your caching model would want to be very different if you know you can cache everything.
Parent
Re: (Score:3, Insightful)
Disclaimer: I work in the data warehousing industry.
Re:Well, a step in the right direction (Score:4, Insightful)
It won't be long before the RAM available in a commodity sever is larger than many shops' database.
First law of data: data always expands to fill all available storage.
Second law: doubling your storage only buys you half the extra time you expected.
Final law: no storage is ever enough.
Parent
Re: (Score:3, Interesting)
The review is slashdotted at the moment so I can't RTFM, but...
If a velociraptor beat an SSD in boot time, well, something is wrong with their test, or perhaps the bios was waiting on the SSD to initialize (entirely possible based on the added intelligence on their controller chipset). I just went from an SLC SSD to a velociraptor and the difference is painful. Boot time is slower. The system is just 'laggier'.
You can't judge the differences between SSD and HDD from charts and graphs on review sites. Re
Re: (Score:3, Funny)
Plus RAID-0 ain't all it's cracked up to be. I had a Dell XPS600 with RAID 0 and one of the drives went kaput. Guess what happens to all the other drives then ? They're useless. 4X drives in RAID-0 means you have four times the chance of having a dead weight for a system.
RAID0: Optimised for failure.
Re: (Score:3, Insightful)
Plus RAID-0 ain't all it's cracked up to be. I had a Dell XPS600 with RAID 0 and one of the drives went kaput. Guess what happens to all the other drives then ? They're useless.
RAID-0 is exactly what it's cracked up to be. It just may not have been what you're looking for.
More Details and Benchmarks Here (Score:5, Informative)
The PCMark Vantage tests are especially impressive: http://www.hothardware.com/Articles/Intel-X25M-80GB-SATA-Solid-State-Drive-Intel-Ups-The-Ante/?page=7 [hothardware.com]
Damn it intel (Score:5, Funny)
You were only supposed to blow the bloody doors off!
It's not the speed, it's the storage (Score:5, Insightful)
Re:It's not the speed, it's the storage (Score:5, Insightful)
Or you split up your expectations.
Honestly, how much space do you need for the OS and programs? Have an SSD for these functions, and a traditional HDD for pure space requirements. That'd be more economical too, at least in the short term.
Parent
Re:It's not the speed, it's the storage (Score:4, Interesting)
At current improvement rates, I think that you're looking at 7-10 years before SSD becomes cheaper than 3.5" form factor drives for sheer storage. We seem to have been lagging at around a terabyte for a while. Meanwhile it seems that SSD is doubling in capacity per $ at it's 'sweet spot' each year at the moment.
Going by performance improvements, it'll only be a 2-4 years before companies start replacing their platters with solid state for intensive database operations, especially those biased towards reads. Those 10k-15k RPM drives are significantly more expensive and store less than 7200/5400 RPM drives.
The article mentions $595. Looking up, a 300GB 15k HD is $400 for an OEM. That's 5 times the size of the 80GB SSD mentioned in the article. Figure on a doubling each year, that'd be 3 years before the SSD exceeds current models. Figure in the lower power requirements and such, and I can see SSDs selling well before reaching parity based purely on size - their improved seek time, lower power demands, etc...
Parent
More details and Benchmarks here (Score:4, Informative)
Benchmarks start here: http://www.hothardware.com/Articles/Intel-X25M-80GB-SATA-Solid-State-Drive-Intel-Ups-The-Ante/?page=4 [hothardware.com]
One test they never run - FRAGMENTATION (Score:3, Interesting)
Since SSD don't really have "sectors", do they fragment files the same way as HDD?
Also, what would the defrag speeds be?
Re:One test they never run - FRAGMENTATION (Score:5, Informative)
Parent
Re:One test they never run - FRAGMENTATION (Score:5, Informative)
Yes, it would wear the disk out faster, but your original premise is flawed.
Clustering locations would allow for accessing large chunks of data with one fetch, instead of lots of little fetches. If you're old enough, think back to the Blitter on the Amiga and moving contiguous chunks of memory as opposed fragmented blocks.
Remember, RAM can get fragmented just as badly as a hard drive.
Parent
Commercial uses don't fragment (Score:4, Interesting)
What would be interesting would be to put an Oracle database block interface on these puppies, instead of the normal filesystem interface. then you'd just have the database say to the storage "get me block X" and it appears. No filesystem overheads - which given the speed of these things could turn out to be significant.
Looks like we'll be back on RAW "disks" for databases. Plus ca change!
Parent
Re:One test they never run - FRAGMENTATION (Score:5, Informative)
A good SSD has wear-leveling and write-combining techniques that keep the SSD "defragmented" automatically.
And it doesn't matter if the FS clusters are far apart as long as they are close to the SSD's hardware cluster sizes or the SSD intelligently combines them (which is what I believe Intel is doing since they claim a write amplification of only 1.1).
It's possible that the Samsung SLC chip stores data for the wear-leveling and write-combining operations which would remap the MLC in a non-fragmented way.
BTW, let me give you a naive wear-leveling / write-combining algorithm. I'm sure Intel has a better one because they've invested millions of dollars of research and the one I'm about to present to you could be done by a CS101 student:
1) You have a bit more than 80GB free for an 80GB drive (extra memory to take care of bad sectors just like a normal hard drive plus a small amount of required for the wear-leveling / writecombining)
2) You treat most of the storage as a ring buffer that consists of blocks on two levels: the native block size and a subblock size. The remaining storage (or alternate storage which may be the Samsung SLC chip on the MLC drives) is used to journal your writes and wear-leveling.
3) You combine all writes aligned to the subblock size into a native block and write them out to the next free native block in the ring buffer and keep a counter for the write to the block. If you run into a used block, and increment a counter (for wear levelling) and if the counter is below a certain value, you skip it to the next free block, otherwise you move the used block (which has been stagnent) to a more frequently writtento free block (which will now take less of a burden since it's had a stagnant block moved into it).
4) Anytime you make a write, the new sectors are updated in the memory area used for journaling / wear-level / sector remapping.
Assuming your reads can be done fairly quickly at the subblock level, it never matters if you have to "seek" for the reads and the drive won't fragment on writes because they are combined into native block sizes.
Parent
Re: (Score:3, Interesting)
I don't mean to attack directly, but you seem to be just well informed enough to be dangerous. First, you seem to think a quick reboot is something that should be no big deal and happen rather often. This is kind of appalling. If you need to reboot a computer often (more than to install new hardware), something serious is wrong with it or it's OS.
Secondly, this phrase, "Hard drives need explicit defragmenting" is misleading as all hell. Hard drives do not need defragmenting. They're made of platters, he
Re:One test they never run - FRAGMENTATION (Score:4, Informative)
You can't grow a file in the middle. There don't exists any filesystem call that can do that.
Fragmentation only happens if you append to a file, but that kind of fragmentation should not be a problem for ssd, because all blocks(Except the last) will be full, and ssd don't read the 'next' block, any faster then any other black.
Parent
Real use for SSD (Score:3, Insightful)
Western Digital blah blah, 2.5" mobile blah blah. How do they compare to the mainline Hitachi and Seagate 15k Fibre Channel? EMC's SSD offerings? I want to know what I can expect for data warehousing on Oracle RAC.
SSD on PS3? (Score:5, Interesting)
Price is over-rated (Score:5, Interesting)
I ended up buying a refurb Dell laptop for around $1000 with a 64 gig SSD. Was it the latest and greatest? Nope. But it was about $150/200 more than a similarly priced computer with a traditional drive (which of course, was larger). Since the only significant problems I've ever had with my two prior Dell laptops (admittedly a small sample) involved the hard drive, going with the SSD (especially when you include the "cool" factors -- both temperature and nerd-ism) was an easy decision.
But the point is that as SSDs become more prevalent, they become available at cheaper prices. I'm sure that as the Intel drives are rolled out, the "obsolete" drives currently on the market will continue to fall in price and become available to bottom-dwelling cheap-o-s like me who may not be able to justify $1000, but can rationalize $200 without a whole lot of difficulty.
Re:Your SSDs (Score:5, Funny)
Those're STDs.
Parent
Re:Your SSDs (Score:5, Funny)
Those're STDs.
It burns when I read/write
Parent
Re: (Score:3, Funny)
to run vista, or do you need a RAID array of these drives.
Vista does a lot better with slow hard drives than XP or most other operating systems, thanks to superfetch or whatever silly name they give to the precache of apps.
Re:but is it fast enough (Score:5, Informative)
That depends entirely on what kind of RAID [wikipedia.org] we're talking about...
Parent
Re:but is it fast enough (Score:5, Informative)
Parent
Re:Blows doors off? I call bullshit. (Score:5, Insightful)
If anyone's seen the results, it's in first place in speed but not in a "door blowing manner". It's just slightly faster than the next guy.
Pardon me, but it is "blowing down the doors" (and the house too) in some tests, like this one [techreport.com]. More than 3x the number of transactions of the second fastest flash drive? 7x faster than the slowest SSD drive? And the traditional HDDs are so crushed at the bottom I can't make out a ratio, but 30x or more? That is just ownage of the highest level. Yes, the write speeds aren't exactly compelling but for IO and read-heavy uses it's completely mindblowing.
Parent
Re:Blows doors off? I call bullshit. (Score:5, Funny)
Pardon me, but it is "blowing down the doors" (and the house too)
Yes, the write speeds aren't exactly compelling but for IO and read-heavy uses it's completely mindblowing
Great, first the doors, then the house and now your mind...
I guess if there's anything we've learned is this drive really blows.
Parent
Re:Blows doors off? I call bullshit. (Score:5, Insightful)
If you read the article, NCQ actually makes sense. The Intel drive actually finishes requests before the CPU gets around to asking "are you done yet?". That time between the drive finishing and the drive being told what to do next is spent idle. By supporting NCQ, the drive can convince the CPU to send large batches of commands and get rid of that latency.
It's faster for the same reason that FTP is faster than IRC DCC. FTP just keep sending bytes as long as the other end doesn't close the connection. IRC DCC sends a packet, waits for a reply, sends the next packet, and so on.
Parent
Re:where is the (Score:5, Funny)
Probably right next to the dlsyexia tag.
Parent
Re:where is the (Score:5, Funny)
The preferred spelling is lysdexia.
Parent
Re: (Score:3, Informative)
Write rates aren't THAT impressive, good but meh.
Less heat depends on the device, I've seen plenty of HOT SSDs, presumably due to the density of silicon in them and being first generation devices
Better power consumption ... where? Every SSD I've seen doesn't have a power saving mode, in power saving mode, as a general rule, mechanical drives are less hungry than SSDs.
They are really only compelling if you need fast seek times or for use in a laptop where shock (head strikes) is a potential issue at this p
Re:Gonna Take a Little While Yet (Score:4, Informative)
Here's my concern in a nutshell:
Assuming a degenerate workload, with a naive algorithm that never remaps existing data except when it is written, death is swift. Assume a 256 KB flash block. Assume a 4 GB flash device with 2% spare. Assume 70 MB/sec. transfer rate. Assume TCQ/NCQ so that you can queue up requests without waiting for the previous request to complete. At 2%, you have about 81.92 MB of spares, or about 328 spares. You have to erase a block containing 256KB at once (one entire flash block). Write random data on a single data block over and over without caching. At 70 MB/sec. divided by a 256 KB block, you can write 280 blocks per second. That comes to about 1.17 seconds to go through all of the spares once. With a 10,000 erasure limit, that means you destroy all the spares in 2.38 hours. At that point, no further writes can occur because erasing and rewriting a block in place is inherently unsafe. Obviously for a 60 GB disk, multiply the numbers by 15. Even with 100,000 cycle flash, one could kill a drive with a naive algorithm in about four months. Okay, so it wouldn't be quite that fast because you'd have to issue write cache flush instructions between each write, but you're in the ballpark.
On the flip side, with a typical workload, a drive would likely last several years even with such a naive algorithm. This is why I'm concerned. It is quite possible for a company to implement a remarkably naive wear leveling algorithm and mostly get away with it except for a few unlucky people who end up with data loss. We saw this in the HD industry not too long ago with IBM claiming after the fact that their drives were not designed for continuous use. With such a history of reliability corner-cutting from storage vendors, I think there's good reason to expect better transparency from the flash drive vendors about how they are doing wear leveling, particularly if these products are expected to be used in enterprise installations as this drive supposedly is. Fool me once and all that....
I won't even get into the question of how one can possibly achieve anything approaching a 1.1 write amplification rate short of building custom flash chips that allow per-page erasure.... Maybe for certain synthetic workloads, but not for a degenerate workload (e.g. write blocks sequentially with a stride length of the same size as (or larger than) the physical flash block size until you exceed the capacity of the write cache, rinse, repeat).... Otherwise, that seems at least an order of magnitude lower than is plausible. I'd have to see white papers explaining exactly how they're doing this miraculously good wear leveling before I'd trust any low-cycle-count SSDs in anything resembling a production server....
Parent