Go to:

IT Jobs

If the world doesn't yet need the performance of the forthcoming SATA 6Gb/s standard, it soon will, drive giant Seagate will set out to prove in a first ever public demo for the technology this week.

Along with chip partner AMD, Seagate hopes the demo will show off the immense data-shifting potential of SATA (Serial ATA) 6Gb/s, also known as ‘SATA revision 3.0', which as its name suggests, neatly doubles the theoretical throughput of today's 3Gb/s SATA 2.0.

In real world conditions, this doubling should add up to a hard drive read out of buffer rate of 589MB/s, the company claims, compared to SATA 2.0's peak of 288MB/s.

But do PC users need this sort of performance given that today's best hard drives struggle to saturate even SATA 2.0's capabilities?

"We always want data rates on the bus to be faster than the disk," says Marc Noblitt, Seagate's senior marketing I/O development manager. "It is not needed now, but what is needed now is the infrastructure."

In development for three years, getting SATA 6Gb/s working has meant AMD developing a completely new chipset for inclusion on motherboards, working in tandem with an ASIC upgrade for Seagate's drives, he said. That sort of development cycle meant anticipating the application demands and data capacities years ahead.

Capacities play a big part in this reasoning, as Seagate's own analysis predicts that approaching 20 percent of PC systems will ship with hard drive capacities of 1 terabyte or more by 2010. Higher capacities make are, in turn, driven by larger file sizes, which add up to larger amounts of data being shifted on and off disks, all putting strain on today's interface standards.

At the time of SATA 2.0's appearance in 2005, nearly 60 percent of PC hard drives ran to 120GB or fewer.

According to Noblitt, SATA 6Gb/s would also herald better power management and data streaming capabilities, both intended to satisfy new demands being placed on drives.

The first systems shipping with SATA 6Gb/s should start appearing at the end of 2009, though the technology won't become mainstream until late 2010 or 2011, he says.