Print 41 comment(s) - last by Spivonious.. on Mar 22 at 2:31 PM

The new SATA Slimline connector interface
New small-form factor and mobile enhancements specified in revision 2.6

The Serial ATA International Organization, also known as SATA-IO, today announced its latest Serial ATA specification – SATA revision 2.6 (PDF). The latest SATA revision 2.6 adds new physical and performance features to the previous SATA revision 2.5, also referred to as SATA 3.0Gbps, specification.

New physical features added to the SATA revision 2.6 include internal cable specifications for small-form-factor systems and mobile PC applications. SATA revision 2.6 specifies a new internal slim-line cable and connector. The new Slimline cable and connector target tightly packed small-form-factor systems. SATA-IO also defines the internal micro SATA connector with ultra mobile PCs in mind. The new internal micro SATA connector specification is for 1.8” hard drives only right now.

In addition to the small-form-factor and ultra mobile PC benefits of SATA revision 2.6, SATA-IO defines a new mini SATA multilane cable and connector. The new mini SATA multilane cable and connector specification is for internal and external SATA usage. Do not expect to see mini SATA multilane connectors on consumer desktop systems, however. SATA-IO defines the specification for use internally in high-bandwidth backplanes and externally for high-bandwidth external storage enclosures.

Physical enhancements aside, SATA revision 2.6 introduces new performance improvements to the SATA specification. Native Command Queuing, or NCQ, receives upgrades beneficial to desktop and notebook systems. NCQ priority enhancement is the latest feature and prioritizes data during complex workloads. On the mobile improvements sides of things, a new NCQ unload enhancement feature makes it way into the SATA specification. This new feature increases mobile SATA hard drive robustness, especially in drop-prone environments.

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

RE: This doesn't make sense
By pepperxn on 3/7/2007 8:42:03 PM , Rating: 2
Here's the only thing that I could think of:

serialATA is 150MBps or 1.5Gbps
serialATA 2(2.0) is really serialATA-IO. This is spec 2.0. IO doesn't have to be 3.0Gbps or 300MBps. This is what most people don't understand. So before you buy a motherboard, and/or hard drive check the specs.
So spec 2.5 and 2.6 adds some features like the actual speed increase, NCQ, etc. Not all 2.0 drives include the speed boost to 300MBps/3.0Gbps.
There are revisions to the specs like adding new features.
Like within a year or so there will be serialATA at 600MBps, but they'll probably call it serialATA 6 (6.0). This would be spec 3.0.
By owning a 3.0 drive now, you probably really mean 3.0Gbps, which is the 300MBps serialATA-IO (serialATA-2.0)

RE: This doesn't make sense
By MPE on 3/7/2007 11:52:38 PM , Rating: 2
serialATA is 150MBps or 1.5Gbps [snip] you probably really mean 3.0Gbps, which is the 300MBps

1.5 Gb/s is not 150 MB/s. It is actually 187.5 MB/s. Nor 3.0 Gb/s equals 300 MB/s. It is actually 375 MB/s.

RE: This doesn't make sense
By InsaneScientist on 3/8/2007 12:48:48 AM , Rating: 2
Actually in this case it is.

The SATA specification transfers 10 bits per byte for some reason, though only 8 are needed.
Maybe a start and stop bit? I don't have a clue.
All I know is with SATA 1.5Gbps does equal 150MB/s and 3Gbps really does equal 300MB/s

Why can't they keep things consistent? >_<

RE: This doesn't make sense
By AndreasM on 3/8/2007 11:21:52 AM , Rating: 1
It's to reduce interference. However, those 10 bits still only contain 8 bits of real data, so technically you're both right. :)

Wikipedia has more if you're interested:

RE: This doesn't make sense
By Oregonian2 on 3/8/2007 7:14:58 PM , Rating: 2
1000BASE-X gigabit ethernet (over fiber) uses 8B/10B as well, so the gigabit serial stream really runs at 1.25 Gb. It gives DC balance as well as provides a minimum number of data edges from which to do clock recovery (else a large block of data that's all ones or all zeros would be a problem).

RE: This doesn't make sense
By Golgatha777 on 3/8/2007 11:26:25 AM , Rating: 2

Ah, learn something new every day. SATA uses 8B/10B encoding, which means that SATA150 is really 150MB/sec actual data throughput due to the 8bit to 10bit encoding.

RE: This doesn't make sense
By mino on 3/8/2007 12:04:50 PM , Rating: 3
AFAIK those 2 bits are for ECC stuff.
1.5Gbps is physical speed while 150MBps is real data bandwith. Kinda.

RE: This doesn't make sense
By Golgatha777 on 3/8/2007 11:17:54 AM , Rating: 2
Thank goodness someone knows the math.

8bits = 1byte

8Gb = 1GB

3.0Gb/sec = 0.375GB/sec or 375MB/sec (10^3 conversion factor)

This is why many people say there is no performance difference between SATA150 and SATA300 because most hard drives have trouble getting near the 100MB/sec sustained transfer rate threshold, much less 187MB/sec or 375MB/sec.

RE: This doesn't make sense
By johnsonx on 3/9/2007 12:21:30 PM , Rating: 2
Thank goodness someone knows the math.

You're right, someone does... just not you.

RE: This doesn't make sense
By drebo on 3/8/2007 12:55:21 AM , Rating: 3
Here's the only thing I could think of:

You're all idiots.

SATA-IO is the organization, the tech group, that publishes and devises the SATA standards.

There are currently two SATA standards:

The original SATA standard, known as SATA 1 or now SATA 150 or SATA 1.5, referring to the speed(1.5 Gbps).

There was a modified SATA standard that, for a time, was known as SATA II. The speed of this standard was 3.0 Gbps. I believe this was the SATA 2.0 specification, and certain things like NCQ were added when it was upgraded to the SATA 2.5 spec, which is what current drives are manufactured. SATA II drives are no longer called SATA II drives because of a law suit which found the SATA II label was a trademark infringement. Now, SATA II is known as SATA 3.0. It does not mean that it is the SATA 3.0 specification. It means that it is SATA with a 3.0 Gbps bus speed.

For reference, the interface is not that much faster. UDMA 133 had a speed of 1.33 Gbps. No hardware, not even 10K RPM drives, can read that fast, so all that extra bandwidth is lost, though. That's why SCSI drives have never matured past 320. Simply no need to expand the bus that wide.

The next SATA standard will be a 6.0 Gbps standard.

Really, read the Wikipedia article or something. Sheesh.

RE: This doesn't make sense
By johnsonx on 3/9/2007 12:22:50 PM , Rating: 2
I'd mod you up, but I talk too much so I never get to vote. Well said.

"We are going to continue to work with them to make sure they understand the reality of the Internet.  A lot of these people don't have Ph.Ds, and they don't have a degree in computer science." -- RIM co-CEO Michael Lazaridis

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki