Some facts: -Any OS uses one (or more) page size for all its io -A RAID array is build using a mix of a "stripe size" on each of the drives participating in the array -TRIMing less than a stripe is a very difficult task even for RAID 0
Example with any Windows 7 installed on a RAID 0 array build with 2x HDD and a 64KB stripe ==> The OS cluster size is usually 4KB, so you may (simplified process): -Write a small file on the first clusters (#0 = 4KB) -The RAID controller read the first 64KB stripe of the first HDD and rewrite it with 4KB updated -Write another small file on the 2d clusters (#1 = 4KB) -The RAID controller read the first 64KB stripe of the first HDD and rewrite it with 4KB updated -Delete the first file will send a TRIM command to #0 ==> Now, if the SSD TRIM this #0, what would be the 4KB values read from the first 64KB stripe for my 3rd small file to be written ?
This implies that the controller should: -be able to read a dedicated special value for a TRIMmed sector -be able to ignore writing those sectors while updating a RAID 0/1 stripe -be able to compute parity on those sectors
==> And please keep in mind that TRIMing is not a mandatory feature, it just helps the SSD Garbage collector in its erase work...most SSD has some reserved FREE space that makes this TRIM feature as an obsolete one
I'm Running two individual X25-M's G2 versions separately; no raid 0 :( . I really am looking forward to trim adaptation in all the popular Operating Systems.
I'm also a little disappointed encryption hasn't been tested at all. SSDs make a lot of sense in laptops, and it's in those setups you're likely to see full disk encryption.
SandForce has built in encryption but I haven't seen anything about that anywhere.
OCZ's pro models which cost about twice as much as the regular Vertex 2 drives support it with their own software. Sandforce's implementation is not OPAL spec.
Bitlocker will give you 60+% performance. Hardware encrypted drives perform at full speed. The encryption runs all the time. Only the authentication gets turned on or off for supporting it.
Won't encryption be processor limited as opposed to depending on the SSD? Maybe you should try using Westmere (Core i5) as they have hardware encryption support and will be much faster. (I think Windows 7 Bitlocker has support for it? Not sure.) The desktop Core i5 6xx and mobile Core i5 (except 450 and 430) and i7 980X have AES-ni support.
I'm using TrueCrypt 7.0a with a Core i5-520 (hardware AES support) on an Intel 80GB G2 and I don't notice a measurable drop in performance. If I have time later, I'll benchmark it with hardware AES support enabled and disabled to see what the real numbers are.
Current pricing seems to be under $2-3/GB for most drives of the best variety(SF, Indilix, Intel). If this means "roughly twice the capacity at the same price", like Anand says, then we're talking $1-$1.50/GB once prices stabilize. Simple math means a possible price between $600 and $900 for the 600GB drive.
Let's say it comes out an MSRP of $999 just for the hell of it... it won't take long for the price to drop. Considering that the Intel refresh won't support SATA-6Gbps (judging by the performance specs), I can't say I would want one anyway. The next batch of SSDs from OCZ, Corsair, Mushkin, G.Skill, Patriot, etc. will likely all support SATA-6Gbps and blow the doors off Intel just as the current Crucial RealSSD C300 does for roughly the same cost/GB.
Intel would have to be really competetive on price for its refresh to compete against the coming lineup of new controllers from other manufacturers and it doesn't look like Intel has upped their write speeds much...
Ah, but then it occurs to me that Intel's process advantage would allow them to sell higher capacity drives for less... so I guess what I'm saying after all this is that I can't wait to see an SSD wrap-up next February! :)
Plus, did anyone else notice the fact that the Intel refresh drives use 2x the load power??
The G3 Intel SSDs, the refresh, are still SATA-3Gbps. As I stated above, I think this will put Intel at a disadvantage as more next-gen SSDs using SATA-6Gbps roll out about the same time.
"The SATA implementation has been updated to support ATA8-ACS so it’s possible we may see official 6Gbps support once Intel has a chipset with native support."
My whole point is the argument about not having 6Gbps atm really doesn't matter, both from the above statement and jonup's point below
But that will not have an effect on the average user. 250MB/s sequential is more then enough unless you bench freak. I think the weakest point of the G2 and the G1 for that matter was the sequential write and it appears the G3 is addressing the issue. I also believe that the G3 will be superior in random writes at the time of the lunch. As for the sequential read, anyone that is not satisfied with 250MB/s is more than likely running a RAID array so the cost per GB will be more important than the max sequential read per hard drive. At least these are my 2 cents.
Most people won't notice, you're right, but some people do. But that's not really the point, is it? Why should anyone settle for the low end or middle of the road?
Ever use multiple 50Mbps 1080p streams in Premeire Pro while mixing multi-track audio? Throughput is king. The more I can get, the happier I'll be, especially in a single-drive scenario for when I'm mobile.
So you have multiple 50Mbps streams? To saturate 250MBps, you'd need roughly 40 of those running simultaneously. Maybe that's a realistic scenario. Audio doesn't use nearly as much bandwidth as that. Even a 96KHz 24bit audio stream is only about 300 KB/s. Mixing 100 of those together is only roughly 30 MB/s. I suppose that limits your total number of multiple 50Mbps streams to only 32 video streams (with leftover bandwidth).
I suppose you also have the writing to take care of, so you could halve the above numbers (50 audio tracks plus 16 video streams). Is that too few? I don't work in the music or video processing business, but based on what you threw out there, and some simple math, the 250 MB/s is more than adequate for the scenario.
I suppose that's why we have people filling up 93 Octane Gas for their cars that don't require it - they don't mind throwing their money away.
I noticed the power doubling but figure that is the max power for the largest drive (600 GB). I assume the power for the smaller drives should be about the same as original G2's.
The speed went up, and so did the power use. It's been the same case with many SSDs.
Flash chips use hardly anything. The amount of flash chips may not actually be higher for the 600GB drive, but even when it's less, like the current 40GB drives, the change in power use is negligible.
What bothers me is that everyone is still marketing SSDs as "more power efficient," but some of them actually use more than HDDs because they're pretty much brute forcing it.
Where are the SSDs for laptops that actually increase battery life? That would have been possible even before SSDs started blowing HDDs out of the water in sequential speeds.
Not really. They can have all the capacity they want, but as long as the cost/gigabytes is still prohibitively high mechanical hard drives will remain.
True, but for that to make sense, the SMALLER capacity SSDs would need to have a higher over-provisioning percentage than the larger SSDs. Just looking at the capacities (300 instead of 320, 600 instead of 640), it seems likely that the larger SSDs have higher percentage over-provisioning.
But as you say, there is no clear indication from Intel.
X25-M 80GB can handle only 7.5TB writes??? Well, that means, that my SSD should be dead already. I've already 9.5 TB host writes, SSD seems to work fine, media wearout indicator is at 96%.
Can you explain why random writes lifespan should be any different from "normal" writes lifespan?
Anyway, my SSD is heavily fragmented, because I turned NTFS compression on. 80 GB is a little bit too small and 160GB cost a little bit too much for me. Therefore most of writes on my SSD are random. I also encrypted whole SSD with Truecrypt (that disables TRIM). Still, performance is quite good.
Random 4KB writes will have much higher write amplification than larger block sequential writes. Higher write amplification means that more flash memory gets written compared to the actual user data that is written. More flash memory being written means a shorter lifetime.
I contacted Intel about this same question. The 7.5TB write threshold refers to the writes per cell. So the 9.5 TB host writes for the entire drive is of little concern, unless you've written to every cell with 7.5 TB of data, then you'd have a problem, but until then, it's nothing to worry about.
Don't forget that SSDs complete writes and reads faster than HDDs. Their peek power consumption is for less time so total power consumption should be reduced for most applications.
The power consumption should be listed per GB so we can compare. But if the max. capacity is increasing by nearly 4x then doubling the max. power required may not be too bad.
Apparently SSDs are not very good at idling so that may be an area where improvements can be made. Surprisingly they are not as cool as many people assume although I have yet to find a clear article on the subject.
In the article you say that the upcoming enterprise drive will be the first to use MLC, but in the chart comparing E-class drives, the 50 nm product is listed as using MLC.
What's the deal with Intel using MLC instead of SLC for the next gen X25-E? I got all excited when Intel first announced the enterprise refresh, but now it's just disappointing. I probably won't ever buy the E drive, but I still wanted to see it blaze through the benches =(
As usual the early adopters have been screwed. By the time mature useful ssds come out the first gen devices will be dead as evidenced by jack.FXs post. That's the price you pay for bleeding edge performance and bragging rights. My 3 striped cuda 7200.12s are still running strong and fast enough for now. This new generation may be the one to finally push me over to the ssd side though. I'm still waiting for that $1 per GB ratio though. A 160gb 3rd gen intel drive for $160 or less would STILL be hard for me to justify but it would be mighty tempting.
Funny... I'm never an early adapter and this time I'm so glad I am. That 30GB SSD I bought a while back put me back a few, but the improvement in DB query time is 10-50 folds compared to a 10K RPM raptor. When have we seen this kind of improvement in GPUs, CPUs, RAM? Maybe in the 90s...
I dont mind being 'screwed', I knew this when I bought it. Its made for speed not longevity and thats what I paid for.
And instead of pushing all the high use stuff like caches onto mechanicals as many suggested, Ive put them all on the SSD instead and still ive only got 1.8TB of host writes used up.
My philosophy is by the time its starting to get full and fail, SSD technology will far exceed todays standards and Ill be looking to upgrade anyway. Have fun with the cudas!
What needs to happen for SDD is a nice drop in price. It's already faster and more convenient than HDD. It's the fact the it cost 20 times more that slows it adoption.
This G3 spec is at best just similar to Sandforce 1222 controller. Which will properly be even cheaper then Intel's controller. 2nd things is the use of very large capacitor for DRAM cache which will bring the cost of SSD up.
The numbers provided are quite intriguing and suggest that the -M drive might actually be a better enterprise drive in many situations. You'd mentioned that "Intel tests 4KB random write performance differently on enterprise vs. consumer drivers so you can’t directly compare the numbers between the X25-M and X25-E." This may be true, but presumably you can compare across generations:
The X25-M and X25-E have similar performance in tests presented here - 4k random read used disk = 35.8 vs. 35.4 MB/s for the X25-M and X25-E respectively. The G3 drive is quoted as offering almost 5x that of the X25-M whereas the X25-E performance only improves by about 50%.
Taken together with the identical sequential and random read speeds and the nearly identical sequential writes (170 vs 200 MB/s), database operations and web servers would seem much better served by the X25-M. Somehow I find that conclusion unbelievable. I just can't believe that the difference in random writes would be enough to tip the balance. I must be missing something.
On the subject of specs, I notice that the Corsair Force is faster or identical in all benchmarks vs. the X25-E but on the AnandTech Storage Bench, the X25-E wins. There is clearly more to the benchmarking story.
A 0.075W power consumption at idle is fantastic, but the 6W at load is a deception, although it makes sense, since it's memory and the more of it there's it, the more power it requires.
Seeing as density is doubled from the second generation to the third generation, it makes sense that the power consumption at load doubles. Then again, most of the time, in my case anyway, the drive will be idle, so this might not be such a bad thing after all.
I'm considering a SSD for my next laptop (Buying sometime next year probably, waiting on AMD/Intel new platforms) and battery life is very important to me, so I'd love to see some battery life testing on a laptop, comparing a 5,400/7,200 rpm HDD, G2 and G3 SSD if possible, when those SSDs come out.
Why would doubling the non-volatile memory necessarily involve a significant power increase? We're not talking about DRAM which needs to get refreshed every few hundred microseconds, nor are we talking about SRAM which has 6 transistors leaking 100% of the time, in use or not. This is all Flash, which will retain state even if you turn the devices off. Doubling the capacity involves increasing the amount of peripheral logic and drivers and such, but that is by no means the dominant factor in active power. Shrinking lithography will decrease the length of wires between features, and smaller devices themselves have less capacitive load, so maintaining one size should even decrease power even acknowledging that smaller transistors may have more off current.
I was expecting this drive to be 6 Gbps and just run at 3 Gbps until intel released their chipset. So this means we will see the refresh of these drives supporting it, not a huge deal but unexpected. I also would have thought write speeds would have been slightly higher tho still not that important for an OS drive.
So it looks like pricing and capacity will be there big advantage with this gen.
Being a current 160GB G2 owner I really don't see a need to upgrade to one unless I have some give me a good price for my used drive.
My usage pattern is also not the same as Jack FX. I've owned my drive since dec 08 and only have like 820GB's of host writes so I think this drive will easily outlast any hard drive that i've owned.
So pro's
Great pricing Capacity increased durability Reliability AES
Cons
No 6 Gbps Sata Writes speeds could have been faster
As SSDs increase in storage capabilities they further encroach on traditional HDDs place in IT. Anand, how much longer until SSDs are affordable for file storage and practicality (>1TB)? Also, what is the lifespan of data on an SSD? Is there any degredation or data loss over time?
I've owned both the Intel X-25M and the SandForce-1222 (Patriot Inferno and Corsair Force series). My experience is the SandForce has some VERY serious bugs, and even if it somehow works properly for you, it's still noticeably slower than the Intel X-25M in actual desktop usage. Sequential transfer speeds are only meaningful when you're transferring huge files. For 98% of your average computer usage for an average user, small (4k) random read/writes are going to have a MUCH bigger impact on "speed" that you can see.
And the bugs I mentioned are very big ones. When I put the SandForce drives in my computer, I was getting BSODs and random crashes constantly. That was on a brand-new Intel server mobo with dual Core i7's, not some cheap MSI board. SandForce doesn't want to admit that there are some major problems with their firmware, and as a result, the vast majority of people buying those drives are now having to deal with problems that make the drive unusable and being given no more info than "a new firmware should be available soon." Pfffft.... for $350, I expect better than that. Haven't had a single problem with my Intel X-25M, not to mention the improvement in performance over the SF-1222 drives. And the cost is much better as well. Again, the Intel X-25M G2 has been, for me at least, SIGNIFICANTLY and noticeably faster than the SandForce drives in every scenario other than copying a HUGE file (such as a DVD iso) to another SSD.
Don't buy into the marketing hype behind the sequential transfer rates that are advertised for the SandForce drives. Do some actual benchmarking on your own and you'll find that the SandForce only reaches those benchmarks in very specific situations which you will rarely, if ever, see on your actual desktop. SandForce's claims about higher speed are a joke to the point of being borderline fraudulent, and the online reviews regarding those drives have opened my eyes as to which reviewers out there can be trusted, and who is just copying and pasting marketing press releases and calling it a review.
Anand earned my respect with some of his coverage on the SandForce drives. This was one of the few sites to point out some of the MAJOR problems with the SandForce controller and to actually look beyond the empty promises of 285MB/s sequential transfer rates. People need to understand the difference between large sequential I/O and small random I/O, and how that impacts desktop performance. If more people knew that less than 2% of your average desktop use will involve large sequential transfers, it might help consumers make more educated purchases.
Also, just picked up some Intel X-25E 64GB drives for some servers at work, and the results have been incredible so far. SLC Nand, FTW!! Performance is on-par with ramdisk for every scenario I've tested: high-load DB handling 3,000 transactions per second, a Zenoss Core install monitoring 5,000 devices (so again, TONS of tiny reads/writes), and our corporate email gateway which is doing a ton of filtering (spamassassin, pyzor, razor, clamav, etc) and handling about 500-1000 messages per second. Those are just a few scenarios where using SSD has given us dramatic performance improvements over HDD, by a signficant order of magnitude. We finally have non-volatile ramdisk!!
linuxpro has a point. I've checked the newegg ratings for Intel SSDs versus the competition, and Intel has the opposition beat for reliability (e.g. number of low egg ratings). Maybe things have changed since I last looked though. I have some x25-V and I love them.
And also, most typical computer use is reading lots of small files, not transferring video back and forth (which is usually stored on HDD because of $/GB anyway).
I have been holding off on a Sandforce drive for that very reason. It looks like it will be Intel for me, though I will wait to see how the G3 version works. I have SSDs on my laptop and 5 other PCs where I don't want to deal with disk crashes. But on my main PC, I still use a platter.
I am not seeing any fast drop down in price untill other manufactures comes with their smaller node models. Intel can collect some money by being the first in this smaller node SSD techology. In longer run allso us consumers will profit for this ofcource, because the production cost will come down with higher dencity. But year 2011 or 2012 may be the year when SSD really may make it's breakthrough. Tha capasity is so big that it is usefull also for storage and not just a boot drive, for the prize.
What is the square area of a device that is 34nm on a side? What is the square area of a device that is 25nm on a side? What is the ratio of the area of the two devices?
Multiply all the above answers by a common area and compare the device counts. Extra credit: Convincing marketing that rounding is not the same as truncating. Extra credit: Comparing the size of the minimum lithography feature with the size of the minimum printed transistor.
While I do not see myself getting a 600 SSD anytime soon, the possibility of their arrival means that the workables at 250 and lower will now be coming within reach.
I still think that this technology is still rather nascent and needs to stabilize, but as they get the hand of it, you will see buisnesses START to buy this and once that happens you will see the pricing models begin to change....
The thing is, what is the best split for installation? Obviously, games that do heavy data loads would not see much from being installed on an SSD (first load time, that's it), but what proggies benefit the most from the boku buckeroos? Is there any way to split the load?
Are you sure the G2 doesn't store any data in the cache? My X25-V loses the last written data if I pull the plug. To me that "safe cache" doesn't mean that the cache itself is new, but that it's actually safe this time. Or maybe the data is just cached somewhere else, but the same thing happened on both AMD and Intel chipsets with both XP and 7, and the mechanical HDDs in the same computer doesn't have this problem (smaller 8MB caches?).
Your OS is likely caching writes. Any data cached in RAM will be lost if power is lost before the data is committed to the SSD. Reason: Even with very large amounts of cache memory on hard drives or SSD's, adding an operating system cache will improve performance. In addition to simply having more data cached, the OS also "knows" more about the data, and how you are using it. To a hard drive, its all just a blob of bytes.
You can turn off unsafe caching by disabling "Advanced Performance" for that drive in your Windows control panel.
"Intel went one step further and delivered 5x what the OEMs requested. Thus Intel will guarantee that you can write 100GB of data to one of its MLC SSDs every day, for the next five years, and your data will remain intact"
5 years is about 1,825 days. 1,825 * 100GB/day = 182,500GB total
That's just over 178 TB. Even the new, unreleased SSD figures of 30TB - 60TB doesn't even come close. In fact, using these figures, my 160GB G2 can only be filled up about 96 times before it dies.
So finally comes the 25nm NANDs! It is a great breakthrough from the 34nm SSDs `cuz if Intel prices it aggresively, we finally get a SSD with around 1$/GB and we will be able to enjoy a 40GB boot-drive at very low price-as cheap as a 500GB HDD we use today. But still I would not get this SSD `cuz my laptop has only one HDD bay. And the premium I have to pay for the 300GB or 600GB model is still too high. And I still can't get a decent capacity from one that I can afford. 0.5$/GB, namely the next generation SSD would be a good choice because with only 160$ I can upgrade to a 320G SSD, that's finally enough for daily use. Seagate's Momentus XT SSH (Solid State Hybrids) would be a great choice during this transition period. It's priced at USD 130 and has 500GB Capacity. Though we can get a 1TB HDD at much lower price, the performance boost it provides is considerable. I've read many tests on Momentus XT and it receives really good comments! I'm not sure how Seagate got such high performance boost with only 4GB Reading Cache so it is really amazing when I see it overrun even the fastest HDD, the WD Velociraptor (well, the Seagate 15000rpm Cheetahs aren't for PCs or laptops lol). Hope the seagate engineers can add a bigger Reading-Writing Cache to make it behave more like an SSD! And I would definitely choose the SSH as the upgrade for my turtle-speed 5400rpm HDD before I can afford an SSD.
I may be making a mistake, BUT let's suppose the maximum 4KB random write speed is 150MB/sec. That means you can write a GB (1024MB) in 6.83 seconds. In 6991 secs (116 mins), you can write out 1 TB. The minimum lifespan of the older drive is 7.5TB, so that means you can use up the life span of the drive in 870 minutes, or roughly 14 hours and 30 minutes. So if you were running a continuous security web cam, or buffering video, or doing some type of application that writes all the time the drive would last less than one day. And these newest versions would only last 2 or 3 days.
However, I read some other research suggesting that most people afraid of wearing out their SSD's needn't be. That they were often rated for over 100,000 cycles or data write at bare minimum. Looking at it this way, a 250GB SSD would take 6 seconds to fill each GB at 170 MB/sec. So 1506 secs or 25 minutes to complete one full write cycle. So 2,509,804 mins or 41,830 hours, or 4.77 years of writing continuously, 24 hours a day. Other estimates put the write cycles upwards towards a million, giving the drive 47 years of 24 hour a day writing. In fact one such article explains another similar drive would last 57 years at it's top speed of 90MB/sec.. Or something roughly like this. It's on the "SSD Myths and Legends" site - http://www.storagesearch.com/ssdmyths-endurance.ht...
Everything I've read before this, seems to indicate at least 100,000 cycles if not more. So I'm confused as to why they are hinting such a low life span on these drives. I'm likely misunderstanding this. Either that or it's a likely a typo. At least I hope so.
It would be extremely difficult to find enough data to continuously write 150MB every second. That security camera would likely feed less then a MB a second to the drive.
Not that your math is off but SSD's are used for their low seek times and speed with random read and writes.
What benefit would you get with having one sitting in a box recording video feeds all day?
The price alone for an SSD would not make that suitable choice for that example which would require a vast amount of storage, archiving Video feeds will not be small.
Your point is valid that most users are better off recording video to hard disk drives. However, SSDs would be great for capturing and editing raw high definition video at 100-200 MB/s. Later the video can be transferred to hard disks for archiving, and the empty SSDs used again. It is conceivable that an application like this would eventually exhaust the life span of the SSDs, but the cost:benefit is justified for a professional production company.
true noblemo. Thankfully, the other article I read seems to think the SSD's can write much longer even if they are writing at their maximum speed 24/7. In fact it thinks they will last many years.
@Makaveli, If you read the article in the link I posted, you'd know that the video feed was only ONE example that would entail a lot of writing to the drive. I have no use for writing video to the drive, but there are tons of other applications that write lots of data. So I think you actually missed the point here. Of course the advantage of the SSD is seek times/random read etc. But the amount of data quoted in this article for the life span is really very small. So low I hink it must be a typo. Just making a mirror image of another HD or moving 200GB back and forth in a server would wear that out. Again, my point was that the longevity is usually thought to be far higher than what this article states. What I said had NOTHING to do with the desire to write video to the drive.
It's really hard to get a sense of the type of performance you can get from ssds in the real world even when using pcmark vantage, especially for games. I would love to see more real gaming tests like was done here. http://www.anandtech.com/show/2614/14
I posted the questions in forum, and asked, and with Anand's articles, come to the basic conclusion is, we have reach the optimum speed gain with SSD.
i.e the bottleneck no longer lies within SSD, it is the CPU who could not decode and operate fast enough , or OS and software are not totally optimized for SSD. At least that is with the case of Sandforce.
The specs that matter for loading applications, game levels etc. and booting the OS (as opposed to copying large files) fast are random IOPS right? And really nothing else?
6.0W max power usage means it passes all common hard drives today. Will this be a problem in future generations of SSDs, that power becomes an issue. Laptops?
6 watts does seem alot, but I believe someone else already touched on this a few pages back. Since the SSD is so much faster than a Hard drive it will not stay at full load as long as a hard drive would. So the shorten amount of time to complete task should balance out the slight higher load over the previous gen.
You're going to have to do better than 170MB/s sustained writes. Current generation drives (and even previous-generation drives) beat that handily; I expect Intel's *next* generation drives to be at least as good as the first-gen Indilinx drives!
You'ld get a lot more out of it just by purly running windows 7 and having the file system be native ntfs. Macbooks are also fairly slow. So when you go to do bootcamp etc, you don't get that great of an end product. Also they have slow core2duo. An i7 based pc tower would be best. If it must be mobile, then an i7 based laptop. You can do that in mac, but they charge double with no real benefits. At least not for me. And I do a lot of stuff, video editing included on pc. Long waits for video to encode is something I don't enjoy on macs. But if you must.. you must :-)
And I'm new to this is that I just been reading the comprehensive article on the reasons why SSD Drives slow down, ever since Windows NT 4 I have allocated the swap file to a separate hard drive to increase performance, it seems that within SSD drive it would be best to have all temporary and swap files on my traditional platter drive rather than using up original speed capabilities of the SSD drive. My question is how do I install so that this happens during setup as if my understanding is correct once the data has been written to the SSD drive the damage has been done so changing these things afterwards and would be less effective than doing it during setup. I'm currently running Windows 7 ultimate 64-bit on Intel x48 chipset with a
they can take the old hdd platters,and start melting the metals from it to make a terminator that breaks into ssd plants,and destroys intel to save hdds
Intel's SSD's are better at wear leveling and least likely to fail prematurely. Random access is top notch, after reading many reviews on multiple sites about OCZ and other brand issues. Sandforce controllers tend to be over-rated and sequential r/w speeds is not the biggest factor in OS drives. I can vouch that a 4yr old processor is booting in 20 seconds usable with Win 7 pro on an 80 GB X2 drive. Thats compared with about 1.5min on a 7200RPM Seagate. There is alot of great benchmarks online that show the Intel G2 series drives still perform extremely well and are the most reliable with a 1.1 wear leveling ratio. If you drive is wasting writes thats lost lifespan and to me reliability is the number one factor. You can check reviews on NewEgg too and see how many DOA's or Dead in a week drives people are dealing with. Ofcourse properly aligning the drives to 4K sector size is critical as well for the SSD's. Make sure if you image one using Acronis or other app that you check alignment. I had to use Paragon Alignment Tool to correct mine. Also on the Intels its important to schedule drive optimizations to take advantage of TRIM. If you do that your SSD should stay lightning fast. For the average Joe an SSD is going to make a much more meaningful difference than a processor upgrade. The hard drive is still the biggest bottleneck on everyone's systems. My older system boots and responds faster than my quad core now. Now about those G3 drives, anyone have an official release date on inside scoop. My thought is Intel may be having issues working with the new 25nm size and reduced writes. Makes you wonder if its worth shrinking the die anymore using the existing tech. Personally I wouldnt mind having a 3.5" 1TB using the current tech since its proven to be a workhorse. Ofcourse the price on such a beast would be way out of my paygrade but maybe in a couple years that will be feasable. For now its SSD for boot drive and good old platter drive for media and data files. Either way I wouldnt stop running your full backups!
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
97 Comments
Back to Article
apriest - Tuesday, October 5, 2010 - link
Yay!!! Now, how about trim with RAID 0? I'm still waiting for that with my G2 SSDs...HachavBanav - Tuesday, October 5, 2010 - link
Some facts:-Any OS uses one (or more) page size for all its io
-A RAID array is build using a mix of a "stripe size" on each of the drives participating in the array
-TRIMing less than a stripe is a very difficult task even for RAID 0
Example with any Windows 7 installed on a RAID 0 array build with 2x HDD and a 64KB stripe
==> The OS cluster size is usually 4KB, so you may (simplified process):
-Write a small file on the first clusters (#0 = 4KB)
-The RAID controller read the first 64KB stripe of the first HDD and rewrite it with 4KB updated
-Write another small file on the 2d clusters (#1 = 4KB)
-The RAID controller read the first 64KB stripe of the first HDD and rewrite it with 4KB updated
-Delete the first file will send a TRIM command to #0
==> Now, if the SSD TRIM this #0, what would be the 4KB values read from the first 64KB stripe for my 3rd small file to be written ?
This implies that the controller should:
-be able to read a dedicated special value for a TRIMmed sector
-be able to ignore writing those sectors while updating a RAID 0/1 stripe
-be able to compute parity on those sectors
==> And please keep in mind that TRIMing is not a mandatory feature, it just helps the SSD Garbage collector in its erase work...most SSD has some reserved FREE space that makes this TRIM feature as an obsolete one
apriest - Wednesday, October 6, 2010 - link
It's Intel... they're smart... :-)Seriously, nice explanation of the complexity of trim on RAID.
MrBrownSound - Tuesday, October 5, 2010 - link
I'm Running two individual X25-M's G2 versions separately; no raid 0 :( . I really am looking forward to trim adaptation in all the popular Operating Systems.JHX - Tuesday, October 5, 2010 - link
Anand, for your next SSD tests, please do some full-disk encryption test and see how it affects performance.I have an Intel G2, and using TrueCrypt full disk encryption, I only get 40-45% of the read and write performance.
Is that something you've looked into?
DesktopMan - Tuesday, October 5, 2010 - link
I'm also a little disappointed encryption hasn't been tested at all. SSDs make a lot of sense in laptops, and it's in those setups you're likely to see full disk encryption.SandForce has built in encryption but I haven't seen anything about that anywhere.
eanazag - Tuesday, October 5, 2010 - link
OCZ's pro models which cost about twice as much as the regular Vertex 2 drives support it with their own software. Sandforce's implementation is not OPAL spec.eanazag - Tuesday, October 5, 2010 - link
Bitlocker will give you 60+% performance. Hardware encrypted drives perform at full speed. The encryption runs all the time. Only the authentication gets turned on or off for supporting it.JHX - Tuesday, October 5, 2010 - link
Nope, I tried Bitlocker, their claims are wrong. The SSD takes just as big of a performance hit as it does with TrueCrypt.ssj4Gogeta - Wednesday, October 6, 2010 - link
Won't encryption be processor limited as opposed to depending on the SSD?Maybe you should try using Westmere (Core i5) as they have hardware encryption support and will be much faster. (I think Windows 7 Bitlocker has support for it? Not sure.)
The desktop Core i5 6xx and mobile Core i5 (except 450 and 430) and i7 980X have AES-ni support.
apriest - Wednesday, October 6, 2010 - link
I'm using TrueCrypt 7.0a with a Core i5-520 (hardware AES support) on an Intel 80GB G2 and I don't notice a measurable drop in performance. If I have time later, I'll benchmark it with hardware AES support enabled and disabled to see what the real numbers are.JHX - Wednesday, October 6, 2010 - link
Encryption wears on the SSD a lot more in general.iamkyle - Tuesday, October 5, 2010 - link
So what's going on with 64GB @ 2PM? :PAnand Lal Shimpi - Tuesday, October 5, 2010 - link
:-PAkumaX - Tuesday, October 5, 2010 - link
:-PPerisphetic - Tuesday, October 5, 2010 - link
600GB?!?!?!?!A chill went down Seagate's, WD's & Hitachi's backs...
Seriously, death to the mechanical dinosaurs of yesteryear. They should go the way of the VHS.
aegisofrime - Tuesday, October 5, 2010 - link
Sadly they will probably cost a thousand at launch. Oh well. Hopefully this brings the 80GB ones into the realm of mass affordability at least.therealnickdanger - Tuesday, October 5, 2010 - link
Current pricing seems to be under $2-3/GB for most drives of the best variety(SF, Indilix, Intel). If this means "roughly twice the capacity at the same price", like Anand says, then we're talking $1-$1.50/GB once prices stabilize. Simple math means a possible price between $600 and $900 for the 600GB drive.Let's say it comes out an MSRP of $999 just for the hell of it... it won't take long for the price to drop. Considering that the Intel refresh won't support SATA-6Gbps (judging by the performance specs), I can't say I would want one anyway. The next batch of SSDs from OCZ, Corsair, Mushkin, G.Skill, Patriot, etc. will likely all support SATA-6Gbps and blow the doors off Intel just as the current Crucial RealSSD C300 does for roughly the same cost/GB.
Intel would have to be really competetive on price for its refresh to compete against the coming lineup of new controllers from other manufacturers and it doesn't look like Intel has upped their write speeds much...
Ah, but then it occurs to me that Intel's process advantage would allow them to sell higher capacity drives for less... so I guess what I'm saying after all this is that I can't wait to see an SSD wrap-up next February! :)
Plus, did anyone else notice the fact that the Intel refresh drives use 2x the load power??
Iketh - Tuesday, October 5, 2010 - link
the sequential read speeds are clearly limited by 3Gbps sata as implied by the articletherealnickdanger - Tuesday, October 5, 2010 - link
... exactly my point?The G3 Intel SSDs, the refresh, are still SATA-3Gbps. As I stated above, I think this will put Intel at a disadvantage as more next-gen SSDs using SATA-6Gbps roll out about the same time.
Iketh - Tuesday, October 5, 2010 - link
"The SATA implementation has been updated to support ATA8-ACS so it’s possible we may see official 6Gbps support once Intel has a chipset with native support."My whole point is the argument about not having 6Gbps atm really doesn't matter, both from the above statement and jonup's point below
therealnickdanger - Wednesday, October 6, 2010 - link
The sentence before that seems to imply that he was talking about the new motherboards, not the SSDs:"Despite the Q1 release of Intel’s 6-series chipsets, Intel is listing the new X25-M as being 3Gbps SATA only."
Hey, I hope they are 6Gbps, but even so, why release official numbers that are so slow? (by comparison, of course)
jonup - Tuesday, October 5, 2010 - link
But that will not have an effect on the average user. 250MB/s sequential is more then enough unless you bench freak. I think the weakest point of the G2 and the G1 for that matter was the sequential write and it appears the G3 is addressing the issue. I also believe that the G3 will be superior in random writes at the time of the lunch. As for the sequential read, anyone that is not satisfied with 250MB/s is more than likely running a RAID array so the cost per GB will be more important than the max sequential read per hard drive. At least these are my 2 cents.therealnickdanger - Wednesday, October 6, 2010 - link
Most people won't notice, you're right, but some people do. But that's not really the point, is it? Why should anyone settle for the low end or middle of the road?Ever use multiple 50Mbps 1080p streams in Premeire Pro while mixing multi-track audio? Throughput is king. The more I can get, the happier I'll be, especially in a single-drive scenario for when I'm mobile.
erple2 - Wednesday, October 6, 2010 - link
So you have multiple 50Mbps streams? To saturate 250MBps, you'd need roughly 40 of those running simultaneously. Maybe that's a realistic scenario. Audio doesn't use nearly as much bandwidth as that. Even a 96KHz 24bit audio stream is only about 300 KB/s. Mixing 100 of those together is only roughly 30 MB/s. I suppose that limits your total number of multiple 50Mbps streams to only 32 video streams (with leftover bandwidth).I suppose you also have the writing to take care of, so you could halve the above numbers (50 audio tracks plus 16 video streams). Is that too few? I don't work in the music or video processing business, but based on what you threw out there, and some simple math, the 250 MB/s is more than adequate for the scenario.
I suppose that's why we have people filling up 93 Octane Gas for their cars that don't require it - they don't mind throwing their money away.
<shrug>
eanazag - Tuesday, October 5, 2010 - link
I noticed the power doubling but figure that is the max power for the largest drive (600 GB). I assume the power for the smaller drives should be about the same as original G2's.OneArmedScissorB - Tuesday, October 5, 2010 - link
The speed went up, and so did the power use. It's been the same case with many SSDs.Flash chips use hardly anything. The amount of flash chips may not actually be higher for the 600GB drive, but even when it's less, like the current 40GB drives, the change in power use is negligible.
What bothers me is that everyone is still marketing SSDs as "more power efficient," but some of them actually use more than HDDs because they're pretty much brute forcing it.
Where are the SSDs for laptops that actually increase battery life? That would have been possible even before SSDs started blowing HDDs out of the water in sequential speeds.
tipoo - Tuesday, October 5, 2010 - link
Not really. They can have all the capacity they want, but as long as the cost/gigabytes is still prohibitively high mechanical hard drives will remain.B3an - Tuesday, October 5, 2010 - link
Dont be stupid. You can now get 3TB HDD's. And you can bet these 600GB's SSD's are still going to cost more than any 2TB or 3TB HDD.I want HDD's to die already but it wont happen any time soon.
nitrousoxide - Wednesday, October 6, 2010 - link
That's true. For $900 we can get four 2TB HDDs to build a superfast RAID 0 array :)jwilliams4200 - Tuesday, October 5, 2010 - link
Anand:I would expect the write lifespan to be directly proportional to the SSD capacity.
For example, if you had said 7.5, 15, 30, 60 TB lifespans for the 80, 160, 300, 600GB SSDs, then it would make perfect sense to me.
But this "30TB - 60TB" spec for "80/160/300/600GB" SSDs is hard to believe.
Can you clarify the relationship between lifespan and capacity with Intel?
Anand Lal Shimpi - Tuesday, October 5, 2010 - link
I believe it's not just lifespan but over provisioning that matters as well, the latter we don't have any indication of at this point.Take care,
Anand
jwilliams4200 - Tuesday, October 5, 2010 - link
True, but for that to make sense, the SMALLER capacity SSDs would need to have a higher over-provisioning percentage than the larger SSDs. Just looking at the capacities (300 instead of 320, 600 instead of 640), it seems likely that the larger SSDs have higher percentage over-provisioning.But as you say, there is no clear indication from Intel.
Jack.FX - Tuesday, October 5, 2010 - link
X25-M 80GB can handle only 7.5TB writes??? Well, that means, that my SSD should be dead already. I've already 9.5 TB host writes, SSD seems to work fine, media wearout indicator is at 96%.Anand Lal Shimpi - Tuesday, October 5, 2010 - link
That's the minimum lifespan, you can obviously exceed it :)Take care,
Anand
-=Hulk=- - Tuesday, October 5, 2010 - link
4KB Random Writes (Drive Lifespan)-> 4KB writes!
Jack.FX - Tuesday, October 5, 2010 - link
Can you explain why random writes lifespan should be any different from "normal" writes lifespan?Anyway, my SSD is heavily fragmented, because I turned NTFS compression on. 80 GB is a little bit too small and 160GB cost a little bit too much for me. Therefore most of writes on my SSD are random. I also encrypted whole SSD with Truecrypt (that disables TRIM). Still, performance is quite good.
jwilliams4200 - Tuesday, October 5, 2010 - link
Random 4KB writes will have much higher write amplification than larger block sequential writes. Higher write amplification means that more flash memory gets written compared to the actual user data that is written. More flash memory being written means a shorter lifetime.Alexander Morou - Saturday, February 26, 2011 - link
I contacted Intel about this same question. The 7.5TB write threshold refers to the writes per cell. So the 9.5 TB host writes for the entire drive is of little concern, unless you've written to every cell with 7.5 TB of data, then you'd have a problem, but until then, it's nothing to worry about.XZerg - Tuesday, October 5, 2010 - link
What the heck!?!?!? This is pretty much in line with a mechanical drives - what is the power advantage now? Except for idle consumption:http://www.anandtech.com/show/3734/seagates-moment...
Nihility - Tuesday, October 5, 2010 - link
Don't forget that SSDs complete writes and reads faster than HDDs. Their peek power consumption is for less time so total power consumption should be reduced for most applications.GTVic - Tuesday, October 5, 2010 - link
The power consumption should be listed per GB so we can compare. But if the max. capacity is increasing by nearly 4x then doubling the max. power required may not be too bad.Apparently SSDs are not very good at idling so that may be an area where improvements can be made. Surprisingly they are not as cool as many people assume although I have yet to find a clear article on the subject.
http://silvertonconsulting.com/blog/2010/09/17/spc...
http://benchmarkreviews.com/index.php?option=com_c...
Cat - Tuesday, October 5, 2010 - link
In the article you say that the upcoming enterprise drive will be the first to use MLC, but in the chart comparing E-class drives, the 50 nm product is listed as using MLC.Assimilator87 - Tuesday, October 5, 2010 - link
What's the deal with Intel using MLC instead of SLC for the next gen X25-E? I got all excited when Intel first announced the enterprise refresh, but now it's just disappointing. I probably won't ever buy the E drive, but I still wanted to see it blaze through the benches =(Randomblame - Tuesday, October 5, 2010 - link
As usual the early adopters have been screwed. By the time mature useful ssds come out the first gen devices will be dead as evidenced by jack.FXs post. That's the price you pay for bleeding edge performance and bragging rights. My 3 striped cuda 7200.12s are still running strong and fast enough for now. This new generation may be the one to finally push me over to the ssd side though. I'm still waiting for that $1 per GB ratio though. A 160gb 3rd gen intel drive for $160 or less would STILL be hard for me to justify but it would be mighty tempting.ytoledano1 - Tuesday, October 5, 2010 - link
Funny... I'm never an early adapter and this time I'm so glad I am. That 30GB SSD I bought a while back put me back a few, but the improvement in DB query time is 10-50 folds compared to a 10K RPM raptor. When have we seen this kind of improvement in GPUs, CPUs, RAM? Maybe in the 90s...MadAd - Wednesday, October 6, 2010 - link
I dont mind being 'screwed', I knew this when I bought it. Its made for speed not longevity and thats what I paid for.And instead of pushing all the high use stuff like caches onto mechanicals as many suggested, Ive put them all on the SSD instead and still ive only got 1.8TB of host writes used up.
My philosophy is by the time its starting to get full and fail, SSD technology will far exceed todays standards and Ill be looking to upgrade anyway. Have fun with the cudas!
PsychoPif - Tuesday, October 5, 2010 - link
What needs to happen for SDD is a nice drop in price. It's already faster and more convenient than HDD. It's the fact the it cost 20 times more that slows it adoption.iwodo - Tuesday, October 5, 2010 - link
This G3 spec is at best just similar to Sandforce 1222 controller. Which will properly be even cheaper then Intel's controller.2nd things is the use of very large capacitor for DRAM cache which will bring the cost of SSD up.
So unless Intel aggressively prices these SSD
Chloiber - Tuesday, October 5, 2010 - link
Faster Sequential WriteFaster Sequential Read
Faster Random Read
Faster Random Write
Yeah right.
iwodo - Tuesday, October 5, 2010 - link
Faster Sequential WriteFaster Sequential Read
Faster Random Read
Faster Random Write
Faster then What? IT IS AT BEST SIMILAR TO Sandforce!
It doesn't matter if it is faster then Gen 2, The current King of Performance is Sandforce!
Yeah Right.
Chloiber - Wednesday, October 6, 2010 - link
Faster than Sandforce - in EVERY discipline. I'm talking about real world performance, not completely unrealistic ATTO bench results.PC911mickster - Tuesday, October 5, 2010 - link
Great read Anand!Thanks!
No1uNo - Tuesday, October 5, 2010 - link
The numbers provided are quite intriguing and suggest that the -M drive might actually be a better enterprise drive in many situations. You'd mentioned that "Intel tests 4KB random write performance differently on enterprise vs. consumer drivers so you can’t directly compare the numbers between the X25-M and X25-E." This may be true, but presumably you can compare across generations:The X25-M and X25-E have similar performance in tests presented here - 4k random read used disk = 35.8 vs. 35.4 MB/s for the X25-M and X25-E respectively. The G3 drive is quoted as offering almost 5x that of the X25-M whereas the X25-E performance only improves by about 50%.
Taken together with the identical sequential and random read speeds and the nearly identical sequential writes (170 vs 200 MB/s), database operations and web servers would seem much better served by the X25-M. Somehow I find that conclusion unbelievable. I just can't believe that the difference in random writes would be enough to tip the balance. I must be missing something.
On the subject of specs, I notice that the Corsair Force is faster or identical in all benchmarks vs. the X25-E but on the AnandTech Storage Bench, the X25-E wins. There is clearly more to the benchmarking story.
Mathieu Bourgie - Tuesday, October 5, 2010 - link
A 0.075W power consumption at idle is fantastic, but the 6W at load is a deception, although it makes sense, since it's memory and the more of it there's it, the more power it requires.Seeing as density is doubled from the second generation to the third generation, it makes sense that the power consumption at load doubles. Then again, most of the time, in my case anyway, the drive will be idle, so this might not be such a bad thing after all.
I'm considering a SSD for my next laptop (Buying sometime next year probably, waiting on AMD/Intel new platforms) and battery life is very important to me, so I'd love to see some battery life testing on a laptop, comparing a 5,400/7,200 rpm HDD, G2 and G3 SSD if possible, when those SSDs come out.
chrysrobyn - Wednesday, October 6, 2010 - link
Why would doubling the non-volatile memory necessarily involve a significant power increase? We're not talking about DRAM which needs to get refreshed every few hundred microseconds, nor are we talking about SRAM which has 6 transistors leaking 100% of the time, in use or not. This is all Flash, which will retain state even if you turn the devices off. Doubling the capacity involves increasing the amount of peripheral logic and drivers and such, but that is by no means the dominant factor in active power. Shrinking lithography will decrease the length of wires between features, and smaller devices themselves have less capacitive load, so maintaining one size should even decrease power even acknowledging that smaller transistors may have more off current.Makaveli - Tuesday, October 5, 2010 - link
I was expecting this drive to be 6 Gbps and just run at 3 Gbps until intel released their chipset. So this means we will see the refresh of these drives supporting it, not a huge deal but unexpected. I also would have thought write speeds would have been slightly higher tho still not that important for an OS drive.So it looks like pricing and capacity will be there big advantage with this gen.
Being a current 160GB G2 owner I really don't see a need to upgrade to one unless I have some give me a good price for my used drive.
My usage pattern is also not the same as Jack FX. I've owned my drive since dec 08 and only have like 820GB's of host writes so I think this drive will easily outlast any hard drive that i've owned.
So pro's
Great pricing
Capacity
increased durability
Reliability
AES
Cons
No 6 Gbps Sata
Writes speeds could have been faster
bigboxes - Tuesday, October 5, 2010 - link
As SSDs increase in storage capabilities they further encroach on traditional HDDs place in IT. Anand, how much longer until SSDs are affordable for file storage and practicality (>1TB)? Also, what is the lifespan of data on an SSD? Is there any degredation or data loss over time?Makaveli - Tuesday, October 5, 2010 - link
Owned the drive since December 2009linuxpro7477 - Tuesday, October 5, 2010 - link
I've owned both the Intel X-25M and the SandForce-1222 (Patriot Inferno and Corsair Force series). My experience is the SandForce has some VERY serious bugs, and even if it somehow works properly for you, it's still noticeably slower than the Intel X-25M in actual desktop usage. Sequential transfer speeds are only meaningful when you're transferring huge files. For 98% of your average computer usage for an average user, small (4k) random read/writes are going to have a MUCH bigger impact on "speed" that you can see.And the bugs I mentioned are very big ones. When I put the SandForce drives in my computer, I was getting BSODs and random crashes constantly. That was on a brand-new Intel server mobo with dual Core i7's, not some cheap MSI board. SandForce doesn't want to admit that there are some major problems with their firmware, and as a result, the vast majority of people buying those drives are now having to deal with problems that make the drive unusable and being given no more info than "a new firmware should be available soon." Pfffft.... for $350, I expect better than that. Haven't had a single problem with my Intel X-25M, not to mention the improvement in performance over the SF-1222 drives. And the cost is much better as well. Again, the Intel X-25M G2 has been, for me at least, SIGNIFICANTLY and noticeably faster than the SandForce drives in every scenario other than copying a HUGE file (such as a DVD iso) to another SSD.
Don't buy into the marketing hype behind the sequential transfer rates that are advertised for the SandForce drives. Do some actual benchmarking on your own and you'll find that the SandForce only reaches those benchmarks in very specific situations which you will rarely, if ever, see on your actual desktop. SandForce's claims about higher speed are a joke to the point of being borderline fraudulent, and the online reviews regarding those drives have opened my eyes as to which reviewers out there can be trusted, and who is just copying and pasting marketing press releases and calling it a review.
Anand earned my respect with some of his coverage on the SandForce drives. This was one of the few sites to point out some of the MAJOR problems with the SandForce controller and to actually look beyond the empty promises of 285MB/s sequential transfer rates. People need to understand the difference between large sequential I/O and small random I/O, and how that impacts desktop performance. If more people knew that less than 2% of your average desktop use will involve large sequential transfers, it might help consumers make more educated purchases.
Also, just picked up some Intel X-25E 64GB drives for some servers at work, and the results have been incredible so far. SLC Nand, FTW!! Performance is on-par with ramdisk for every scenario I've tested: high-load DB handling 3,000 transactions per second, a Zenoss Core install monitoring 5,000 devices (so again, TONS of tiny reads/writes), and our corporate email gateway which is doing a ton of filtering (spamassassin, pyzor, razor, clamav, etc) and handling about 500-1000 messages per second. Those are just a few scenarios where using SSD has given us dramatic performance improvements over HDD, by a signficant order of magnitude. We finally have non-volatile ramdisk!!
Finally - Wednesday, October 6, 2010 - link
...and how's the pay @ the Intel marketing division?greenguy - Wednesday, October 6, 2010 - link
linuxpro has a point. I've checked the newegg ratings for Intel SSDs versus the competition, and Intel has the opposition beat for reliability (e.g. number of low egg ratings). Maybe things have changed since I last looked though. I have some x25-V and I love them.And also, most typical computer use is reading lots of small files, not transferring video back and forth (which is usually stored on HDD because of $/GB anyway).
Jim1900 - Thursday, October 7, 2010 - link
I have been holding off on a Sandforce drive for that very reason. It looks like it will be Intel for me, though I will wait to see how the G3 version works. I have SSDs on my laptop and 5 other PCs where I don't want to deal with disk crashes. But on my main PC, I still use a platter.haukionkannel - Tuesday, October 5, 2010 - link
I am not seeing any fast drop down in price untill other manufactures comes with their smaller node models. Intel can collect some money by being the first in this smaller node SSD techology. In longer run allso us consumers will profit for this ofcource, because the production cost will come down with higher dencity.But year 2011 or 2012 may be the year when SSD really may make it's breakthrough. Tha capasity is so big that it is usefull also for storage and not just a boot drive, for the prize.
Axz - Tuesday, October 5, 2010 - link
Can someone explain why a reduction of 9nm (~ 25%) in process would reflect in double capacity for the same price?Thank you.
atkoj - Tuesday, October 5, 2010 - link
Because the components are 2D not linear.32*32 = 1024
25*25 = 625
Double(ish)
mianmian - Tuesday, October 5, 2010 - link
The cost is roughly proportional to area density:75% * 75% = 56%
Iketh - Tuesday, October 5, 2010 - link
the same reason 640x480 resolution is four times the resolution of 320x240... at a quick glance, you'd think it's doublethe same reason why 1080p is double the resolution of 720p
you have 2 dimensions of space, not 1
chrysrobyn - Wednesday, October 6, 2010 - link
What is the square area of a device that is 34nm on a side?What is the square area of a device that is 25nm on a side?
What is the ratio of the area of the two devices?
Multiply all the above answers by a common area and compare the device counts.
Extra credit: Convincing marketing that rounding is not the same as truncating.
Extra credit: Comparing the size of the minimum lithography feature with the size of the minimum printed transistor.
Ninjahedge - Tuesday, October 5, 2010 - link
While I do not see myself getting a 600 SSD anytime soon, the possibility of their arrival means that the workables at 250 and lower will now be coming within reach.I still think that this technology is still rather nascent and needs to stabilize, but as they get the hand of it, you will see buisnesses START to buy this and once that happens you will see the pricing models begin to change....
The thing is, what is the best split for installation? Obviously, games that do heavy data loads would not see much from being installed on an SSD (first load time, that's it), but what proggies benefit the most from the boku buckeroos? Is there any way to split the load?
Chloiber - Tuesday, October 5, 2010 - link
By the way - there is a mistake in the second table. Ephraimbased E-models use SLC, not MLC.ajp_anton - Tuesday, October 5, 2010 - link
Are you sure the G2 doesn't store any data in the cache? My X25-V loses the last written data if I pull the plug. To me that "safe cache" doesn't mean that the cache itself is new, but that it's actually safe this time.Or maybe the data is just cached somewhere else, but the same thing happened on both AMD and Intel chipsets with both XP and 7, and the mechanical HDDs in the same computer doesn't have this problem (smaller 8MB caches?).
Sivar - Tuesday, October 5, 2010 - link
Your OS is likely caching writes.Any data cached in RAM will be lost if power is lost before the data is committed to the SSD.
Reason: Even with very large amounts of cache memory on hard drives or SSD's, adding an operating system cache will improve performance. In addition to simply having more data cached, the OS also "knows" more about the data, and how you are using it. To a hard drive, its all just a blob of bytes.
You can turn off unsafe caching by disabling "Advanced Performance" for that drive in your Windows control panel.
Sivar - Tuesday, October 5, 2010 - link
How long will Intel SSD's last?http://www.anandtech.com/show/2614/4
"Intel went one step further and delivered 5x what the OEMs requested. Thus Intel will guarantee that you can write 100GB of data to one of its MLC SSDs every day, for the next five years, and your data will remain intact"
5 years is about 1,825 days. 1,825 * 100GB/day = 182,500GB total
That's just over 178 TB. Even the new, unreleased SSD figures of 30TB - 60TB doesn't even come close.
In fact, using these figures, my 160GB G2 can only be filled up about 96 times before it dies.
nitrousoxide - Tuesday, October 5, 2010 - link
So finally comes the 25nm NANDs! It is a great breakthrough from the 34nm SSDs `cuz if Intel prices it aggresively, we finally get a SSD with around 1$/GB and we will be able to enjoy a 40GB boot-drive at very low price-as cheap as a 500GB HDD we use today. But still I would not get this SSD `cuz my laptop has only one HDD bay. And the premium I have to pay for the 300GB or 600GB model is still too high. And I still can't get a decent capacity from one that I can afford. 0.5$/GB, namely the next generation SSD would be a good choice because with only 160$ I can upgrade to a 320G SSD, that's finally enough for daily use.Seagate's Momentus XT SSH (Solid State Hybrids) would be a great choice during this transition period. It's priced at USD 130 and has 500GB Capacity. Though we can get a 1TB HDD at much lower price, the performance boost it provides is considerable. I've read many tests on Momentus XT and it receives really good comments! I'm not sure how Seagate got such high performance boost with only 4GB Reading Cache so it is really amazing when I see it overrun even the fastest HDD, the WD Velociraptor (well, the Seagate 15000rpm Cheetahs aren't for PCs or laptops lol). Hope the seagate engineers can add a bigger Reading-Writing Cache to make it behave more like an SSD! And I would definitely choose the SSH as the upgrade for my turtle-speed 5400rpm HDD before I can afford an SSD.
zing99 - Tuesday, October 5, 2010 - link
I may be making a mistake, BUT let's suppose the maximum 4KB random write speed is 150MB/sec. That means you can write a GB (1024MB) in 6.83 seconds. In 6991 secs (116 mins), you can write out 1 TB. The minimum lifespan of the older drive is 7.5TB, so that means you can use up the life span of the drive in 870 minutes, or roughly 14 hours and 30 minutes. So if you were running a continuous security web cam, or buffering video, or doing some type of application that writes all the time the drive would last less than one day. And these newest versions would only last 2 or 3 days.However, I read some other research suggesting that most people afraid of wearing out their SSD's needn't be. That they were often rated for over 100,000 cycles or data write at bare minimum. Looking at it this way, a 250GB SSD would take 6 seconds to fill each GB at 170 MB/sec. So 1506 secs or 25 minutes to complete one full write cycle. So 2,509,804 mins or 41,830 hours, or 4.77 years of writing continuously, 24 hours a day. Other estimates put the write cycles upwards towards a million, giving the drive 47 years of 24 hour a day writing. In fact one such article explains another similar drive would last 57 years at it's top speed of 90MB/sec.. Or something roughly like this. It's on the "SSD Myths and Legends" site - http://www.storagesearch.com/ssdmyths-endurance.ht...
Everything I've read before this, seems to indicate at least 100,000 cycles if not more. So I'm confused as to why they are hinting such a low life span on these drives. I'm likely misunderstanding this. Either that or it's a likely a typo. At least I hope so.
Mr Perfect - Wednesday, October 6, 2010 - link
It would be extremely difficult to find enough data to continuously write 150MB every second. That security camera would likely feed less then a MB a second to the drive.Makaveli - Tuesday, October 5, 2010 - link
Not that your math is off but SSD's are used for their low seek times and speed with random read and writes.What benefit would you get with having one sitting in a box recording video feeds all day?
The price alone for an SSD would not make that suitable choice for that example which would require a vast amount of storage, archiving Video feeds will not be small.
noblemo - Wednesday, October 6, 2010 - link
Your point is valid that most users are better off recording video to hard disk drives. However, SSDs would be great for capturing and editing raw high definition video at 100-200 MB/s. Later the video can be transferred to hard disks for archiving, and the empty SSDs used again. It is conceivable that an application like this would eventually exhaust the life span of the SSDs, but the cost:benefit is justified for a professional production company.zing99 - Saturday, October 23, 2010 - link
true noblemo. Thankfully, the other article I read seems to think the SSD's can write much longer even if they are writing at their maximum speed 24/7. In fact it thinks they will last many years.zing99 - Saturday, October 23, 2010 - link
@Makaveli, If you read the article in the link I posted, you'd know that the video feed was only ONE example that would entail a lot of writing to the drive. I have no use for writing video to the drive, but there are tons of other applications that write lots of data. So I think you actually missed the point here. Of course the advantage of the SSD is seek times/random read etc. But the amount of data quoted in this article for the life span is really very small. So low I hink it must be a typo. Just making a mirror image of another HD or moving 200GB back and forth in a server would wear that out. Again, my point was that the longevity is usually thought to be far higher than what this article states. What I said had NOTHING to do with the desire to write video to the drive.DoktorSleepless - Wednesday, October 6, 2010 - link
It's really hard to get a sense of the type of performance you can get from ssds in the real world even when using pcmark vantage, especially for games. I would love to see more real gaming tests like was done here.http://www.anandtech.com/show/2614/14
iwodo - Wednesday, October 6, 2010 - link
I posted the questions in forum, and asked, and with Anand's articles, come to the basic conclusion is, we have reach the optimum speed gain with SSD.i.e the bottleneck no longer lies within SSD, it is the CPU who could not decode and operate fast enough , or OS and software are not totally optimized for SSD. At least that is with the case of Sandforce.
Pegleg7 - Wednesday, October 6, 2010 - link
What exactly does "data retention" mean in the article? The duration the drive will store data from the point it is completely unpowered?I'm assuming "longevity" corresponds to how many times the drive can be rewritten.
TinyTeeth - Wednesday, October 6, 2010 - link
The specs that matter for loading applications, game levels etc. and booting the OS (as opposed to copying large files) fast are random IOPS right? And really nothing else?Pjotr - Wednesday, October 6, 2010 - link
6.0W max power usage means it passes all common hard drives today. Will this be a problem in future generations of SSDs, that power becomes an issue. Laptops?Makaveli - Wednesday, October 6, 2010 - link
6 watts does seem alot, but I believe someone else already touched on this a few pages back. Since the SSD is so much faster than a Hard drive it will not stay at full load as long as a hard drive would. So the shorten amount of time to complete task should balance out the slight higher load over the previous gen.Guspaz - Wednesday, October 6, 2010 - link
You're going to have to do better than 170MB/s sustained writes. Current generation drives (and even previous-generation drives) beat that handily; I expect Intel's *next* generation drives to be at least as good as the first-gen Indilinx drives!AnnihilatorX - Wednesday, October 6, 2010 - link
Spam above, don't click the link thinking it's an article about SSD trim lolPeanutGallery - Friday, October 8, 2010 - link
I'm planning to buy a 13 inch MacBook Pro soon, and maybe install one of the new Intel SSDs when they come out.Will I be able to use the hardware encryption the SSD? How? Does the usage of BootCamp affect the answer?
zing99 - Saturday, October 23, 2010 - link
You'ld get a lot more out of it just by purly running windows 7 and having the file system be native ntfs. Macbooks are also fairly slow. So when you go to do bootcamp etc, you don't get that great of an end product. Also they have slow core2duo. An i7 based pc tower would be best. If it must be mobile, then an i7 based laptop. You can do that in mac, but they charge double with no real benefits. At least not for me. And I do a lot of stuff, video editing included on pc. Long waits for video to encode is something I don't enjoy on macs. But if you must.. you must :-)o2eazy - Tuesday, November 23, 2010 - link
And I'm new to this is that I just been reading the comprehensive article on the reasons why SSD Drives slow down, ever since Windows NT 4 I have allocated the swap file to a separate hard drive to increase performance, it seems that within SSD drive it would be best to have all temporary and swap files on my traditional platter drive rather than using up original speed capabilities of the SSD drive. My question is how do I install so that this happens during setup as if my understanding is correct once the data has been written to the SSD drive the damage has been done so changing these things afterwards and would be less effective than doing it during setup. I'm currently running Windows 7 ultimate 64-bit on Intel x48 chipset with aCharonPDX - Thursday, December 2, 2010 - link
Spammity spam!Calle2003 - Saturday, January 15, 2011 - link
Any release date out yet?athanasios917 - Saturday, January 22, 2011 - link
they can take the old hdd platters,and start melting the metals from it to make a terminator that breaks into ssd plants,and destroys intel to save hddsCyborg1024 - Wednesday, February 2, 2011 - link
Intel's SSD's are better at wear leveling and least likely to fail prematurely. Random access is top notch, after reading many reviews on multiple sites about OCZ and other brand issues. Sandforce controllers tend to be over-rated and sequential r/w speeds is not the biggest factor in OS drives. I can vouch that a 4yr old processor is booting in 20 seconds usable with Win 7 pro on an 80 GB X2 drive. Thats compared with about 1.5min on a 7200RPM Seagate. There is alot of great benchmarks online that show the Intel G2 series drives still perform extremely well and are the most reliable with a 1.1 wear leveling ratio. If you drive is wasting writes thats lost lifespan and to me reliability is the number one factor. You can check reviews on NewEgg too and see how many DOA's or Dead in a week drives people are dealing with. Ofcourse properly aligning the drives to 4K sector size is critical as well for the SSD's. Make sure if you image one using Acronis or other app that you check alignment. I had to use Paragon Alignment Tool to correct mine. Also on the Intels its important to schedule drive optimizations to take advantage of TRIM. If you do that your SSD should stay lightning fast. For the average Joe an SSD is going to make a much more meaningful difference than a processor upgrade. The hard drive is still the biggest bottleneck on everyone's systems. My older system boots and responds faster than my quad core now. Now about those G3 drives, anyone have an official release date on inside scoop. My thought is Intel may be having issues working with the new 25nm size and reduced writes. Makes you wonder if its worth shrinking the die anymore using the existing tech. Personally I wouldnt mind having a 3.5" 1TB using the current tech since its proven to be a workhorse. Ofcourse the price on such a beast would be way out of my paygrade but maybe in a couple years that will be feasable. For now its SSD for boot drive and good old platter drive for media and data files. Either way I wouldnt stop running your full backups!Edgemil - Friday, February 11, 2011 - link
Does anyone know what the drive´s lifespan were on the Intel G1 drive?Would be nice to see the G1 stats in the comparison table as well.