The Register (fwd)

From: Andre Hedrick (andre@linux-ide.org)
Date: Mon Nov 27 2000 - 20:35:02 EST


http://www.theregister.co.uk/content/2/14907.html

Everyone, do not reply to this.

Because I know that Bill G. has a psuedo name here and I have not figured
it out yet otherwise I would sent this direct.

This is a personal poke at him, that they ingnored the subject until
figuring out that Linux was a member and would have to conform to the
initial SPEC that we had input on and they did not!

The next punch will be to deliver adapter support before them!
Also any other things that are turning quietly in the land of ATA.

Cheers All,

Andre Hedrick
Linux ATA Development


The Register
  28 November 2000
  Update: 00:04 GMT
The Register Biting the hand that feeds IT
Search The Register

Advertisement
AceQuote for hardware
 
The Register
 
Hardware Round-up
 
Advertisement

Linux in, Redmond out of Serial ATA party
Posted: 22/11/2000 at 02:15 GMT

The Serial ATA working group, charged with replacing clumsy parallel ribbons as the mass storage interface inside PCs, says it has completed version 1.0 of the specification. Publication had been mooted for 10 October, but the public version now ought to be available for comment by the end of the year.

Members include APT, which provides intellectual property for storage interfaces, hard disk manufacturers Maxtor, IBM, Seagate and Quantum, and Dell and Intel. The only OS with representation on the group is Linux, and Redmond is a notable absentee - although sources say that it's now very keen to join. When it does, it won't have input to the first specification.

But maybe that isn't such a big deal, as one of the promises of Serial ATA is that it presents the same view to the OS as a parallel ATA drive, and doesn't require special device drivers.

Intel's Jeff Ravencraft, Serial ATA committee chair, told us that any additional effort to tweak the parallel ATA interface beyond the current ATA-100 spec wouldn't be worth the trouble.

"We're not sure if it's doable, and even it was, it would be a very fragile innovation," he said. APT's business development manager Robert Streeby added that there would be reliability benefits for other PC components - particularly CPUs - once the requirement to support 5V I/O was removed.

But the most obvious benefit is that PCs should finally get smaller - or at least in theory. "There's 72 square inches of cabling inside a PC. A thin replacement makes way for better ventilation, and smaller cases."

The first private draft spec is intended to reach 1.5Gbps throughput speeds, although real transfer rates will max at 150Mbps. Further revisions are pegged to raise this to 300 and 600Mbps.

The first spec will support a simple 1 to 1 connections, sources tell us, with one drive to one interface. It's hoped the 1.0 draft will be ratified and made public by the end of the year. Streeby sees devices coming on stream towards the end of next year, which is rather more aggressive a schedule than other will admit to. But with no modifications needed to the software stack, it's down to motherboard support, so it could conceivably squeak in.

And with Apple removing the internal 1394 connectors from its latest G4 towers, Serial ATA should have the volume business to itself. ®

Related Stories
Linux beats Microsoft to support superfast disks
Intel disses 1394 for own technology - again

Sections
Front Page
Hardware
Semiconductors
Software
Networks
Internet
Business
Bootnotes
The Week
Register Merchandise


BOFH 2K: Kit and Caboodle
Flame of the Week
Register Info
Cash Register
Register Full Coverage
Register Links
Hardware Round-up
The Vulture Central Mailbag
The Mac Channel
The Register Forum


If you want daily updates on news, enter your email address below, then click the 'Join List' button.
Powered by ListBot


Story Index Story Index
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Thu Nov 30 2000 - 21:00:18 EST