Slashdot Mirror
Leaked Intel Roadmap Shows 600GB SSD
An anonymous reader writes "Solid State Drives have been trying to fill the mechanical hard drive niche for some time now. The problem is that while flash memory is faster than a spinning platter, it is also much more expensive per gigabyte. Over the weekend details leaked about Intel's SSD roadmap, and what's most interesting about it is that the capacities of Intel's SSDs are going to increase in a big way. First off is a refresh to the high performance X25-M range of SSDs. Currently available in 80GB and 160GB models, these will be replaced by a new design, codenamed Postville, which will come in 160GB, 300GB and 600GB variants."
The price is still far too high. I recognize that an SSD can provide a good performance boost, but still...the prices are way too high. I'll likely give it another year or two before I pull the trigger on one.
Not that any of you care -_-;;
Living With a Nerd
For example, this is a posting using the code name: http://communities.intel.com/message/51359;jsessionid=F3036FCC8C1DD878FCED25A7A6D32547.node6COM
Intel does not have the fastest MLC drives out there (X25-E is SLC), and now they're ditching SLC?
I wonder how their performance will match the other controllers (Sandforce, Indilix, Samsung, etc)... perhaps their new MLC is more along the lines of what Sandforce is doing?
Make sure everyone's vote counts: Verified Voting
I keep hearing people claim reliability issues when SSD articles come along to slashdot.
I have never seen a citation, so I went looking for them via Google but could only find citations attesting to the high reliability of these devices.
Dell's Lionel Menchaca stated in 2008, when it was reported by Avian Securities that Dell was having SSD reliability issues, "Our global reliability data shows that SSD drives [that we shipped] are equal to or better than traditional hard disk drives we've shipped." He further notes that Avian Securities never contacted them and that their numbers were a complete fabrication.
At this point I consider any claims that SSD's are less reliable to simply be a myth derived from dishonest reporting.
Furthermore, there are published studies detailing how unreliable traditional magnetic platter drives are.
Do they have write limits? Yes. Can other parts of the device fail? Yes. Are they more expensive than economy platters? Yes. Is there real world data showing that they are less reliable as claimed? Apparently not.
"His name was James Damore."
People who tune large databases have been IOPS focused for a long time. SSDs enable a new level of IOPS that is about one to two orders of magnitude better than spinning disks. SSDs will allow people to (re)consider all sorts of applications that are currently IOPS bound or IOPS prohibited. Soon Google will be able to keep track of how much milk you have in your fridge, and send you a reminder to buy some when you are near a store that sells it, and have plans to go home afterward so that they can be sure you will be able to refrigerate it.
http://online.wsj.com/article/SB10001424052748704901104575423294099527212.html?mod=WSJ_Opinion_LEADTop
"Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
FWIW I've already seen flash device failures (SSD and USB sticks). They tend to fail into a RO mode rather than a blank, or unreadable mode. This is a good thing from a data integrity standpoint (though a bad thing from an IS standpoint).
I personally would feel comfortable using SSDs in a transaction server and such from a data integrity view, but I'm not sure if they could actually handle massive IOPS for a sustained period. Massive OPS, however, they seem to be awesome at, and that's how we're currently using them. A front end cache for largely static datasets that need high read availability. Where we used to be bottlenecking on the drive, we are now bottlenecking on the controller logic.
whois gawk date unzip strip find touch finger mount join nice man top fsck grep eject more yes exit umount sleep dump
Knowing Intel, most likely the new drives, which double the capacity, will remain the same price. Therefore your 160gb drive for $399 will be 320GB for $399, and imho approaching $1 per GB on SSD is a big freaking deal. Of course this info is an early leak and Intel has no mention of price. But with a new fab and smaller nm, most likely Intel will deliver on this theory.
Flash memory is actually very cheap when you compare it with other memories made of silicon. In term of density (bit/nm^2), there is nothing better today: a flash cell is approx 3 times smaller than a DRAM cell. Additionally, today, a single flash cell can store up to 3 bits (Multi Level Cell, MLC as opposed to Single Level Cell, SLC). However MLC-based flash memories require more time to read and to program pages and have a lower life time.
As the price of a chip roughly depends on its die area, it also makes flash the cheapest memory existing today. Today MLC Flash is ~11x denser than DRAM, while SLC is around 3x denser, that makes it approximately 11 times and 3 times cheaper than DRAM, respectively. Look at the prices of DRAM and Flash in your favorite store, you should get close to these ratios.
In the future prices should lower as we make flash even denser. However there is a limit to this shrinking: in 3-4 years the price reduction of flash should slow down. Another issue I heard about is that if everybody in the world wanted to replace their magnetic disk to SSDs today, there simply would not be enough silicon available...