SATA Revision 3.0 specification to double transfer speeds to 6Gbps
By Darren Murph 
posted August 19th 2008 7:59AM
posted August 19th 2008 7:59AM
While not the most mesmerizing news to flow from 2008's Intel Developer Forum, this is a welcome piece for anyone who appreciates hasty transfers. The Serial ATA International Organization (SATA-IO) has announced that the partially complete SATA Revision 3.0 specification will eventually double data transfer rates to 6Gbps. This week's release gives the green light to developers to begin designing products based on said tech, though the finalized paperwork won't be ready until later this year. On a related note, the same consortium has announced the formal launch of its certified logo program, which will enable consumers to easily recognize when a product has passed SATA-IO interoperability testing and is a bona fide SATA 6Gb/sec-enabled device. Enough chatter -- wares, please?[Via SmallNetBuilder]
Read - SATA 3.0 specification
Read - SATA-IO certification logo
6Gbps doesn't seem very impressive considering USB 3.0 has a transfer speed of 4.8Gbps. Will it be enough to meet the demands of the next generation of SSD devices?
are internal components wired by USB? no. this is a hard disk controller we're talking about here. no bottleneck and misleading max throughput stats.
Useless. Hard drives and SSDs get no where near the limit of SATA2 already. What's the point?
Bragging rights. Go home and tell the wifey about the new spec you worked on and how fast it is, casually leaving out the fact that it's worthless.
I would think that SATA 3 initial aim is for the server market. With entry level servers using starting to use SATA for cost savings, SATA 3 should be able to push them more into the higher end level and compete with the faster SCSI technologies.
A 250 MB/s SSD was recently announced, a couple of years from now we could see even faster drives. There's that iodrive thing that was announced a milion years ago and is now being used in some servers, it has a read speed of 800 MB/s and currently uses the PCI-E x4 slot. Wouldn't it be nice to use it via SATA instead?
Ok Im confused, whats the point of having fast bandwith between components that cannot go that fast. IE: the hard drive, can only sustain what, an average 150mb/sec?
Currently, yes, but give it a couple years and SSD's will be hitting the performance limits of SATA II, and unless you have a faster specification in place before that point, you'll be hit with a problem where you'll have fast drives, but no way of getting the data off of them
Raid setups?
From what I understand SATA only allows 1 device per channel, so RAID setups don't saturate the SATA bus, they work against the controller's bus (PCI-EX or PCI).
And Engadget posted this article a couple days about a company already working on plans for a 600MB/s SSD.
http://www.engadget.com/2008/08/17/indilinx-and-mosaid-aim-to-squeeze-600mb-s-out-of-ssds/
So those kinds of delicious speeds are coming sooner or later.
With that said, as mentioned above, I'm disappointed that it's only 20% faster than USB3. I guess it doesn't matter when nothing really uses that much bandwidth yet, but it just doesn't feel very future proof.
Anyone know what kind of bandwidth a PCI-EX x1 lane has (I assume a x4 has 4 times that and so on)??
Perhaps it uses less power or cpu time? Maybe it is more resistant to noise? Easier to implement.
Designed for hardware concatenation devices?
There are always reasons. Often the higher numbers give non-technical/marketing people a "reason".
Of course there is always the possibility it's entirely PR driven.
Yeah, it's quieter, unlike those super noisy SATA v2 connectors and cables which drive me freaking nutz. But at least they're better than the PATA ones which would keep the whole block up at night.
Uh, Mile, I don't think that's the kind of "noise" James was talking about. See http://en.wikipedia.org/wiki/Electronic_noise and related articles.
As for the merits of the new specification, I agree we won't need it for years, but I guess it's better to have it before it's needed than after it's needed (as with the Wireless-N specification debacle).
Or you could make it use USB 3.0. That would be kinda cool, any harddrive would be portable..
Oh, no you didn't!
I'd like to know just how fast USB3 is going to be, the likelihood that it will run at its theoretical max is very unlikely.
yeah, this would be interesting if they could make harddrives that were actually saturating SATAII. Im all for fast link speeds, but HDD technology has been lagging behind lately. dissapointing
"On a related note, the same consortium has announced the formal launch of its certified logo program, which will enable consumers to easily recognize when a product has passed SATA-IO interoperability testing "
Good. Hopefully that means WDC will make sure to release 100% compatible SATA drives in the future. The SATA-IO specifications also include connector placement, something WDC ignored with the initial release of the Velociraptor but still claimed the drive was a SATA drive.
It would be nice if they supported hot swapping properly as well. Currently the spec allows it but no one implements it on desktop machines.
a S S hole,
giving out useless standard to force use buy expensive new hard drive.
Go spend your resources to push the SATA 2.0 devices to the limit first.
Moron.
o,schmitty
I read your replies on different posts.
You are always try to attacking the repliers. It is such a pity.
Anyway, saying Maroon 5 is good , but you never say the other is a moron because it is personal.
Wooo Hoo! Now my HD DVR can send data twice as fast to my external hard. Of course that will be of little use to me, but twice as fast is good, right? This is one more reason to wait until Q109 to setup my new system though. Having that kind of access speed is sexy. I'm with the other worriers though in the vein that it's doubtful any consumer hard drive will really be able to test this new limit. If it's pushing the SATA 2.0 limit I'll be impressed.
Thanks,
Scott
http://www.topclassactions.com
Great... now if we could just get drives that could even come close to utilizing the speed, then it'd actually be worthwhile.