Seagate and AMD Demo SATA 6.0 Gbps -- Should You Care?
In case you missed today's big announcement, Seagate joined up with AMD and got its world-first achievement for demoing the next-generation SATA specification, SATA 6Gbps. But what does that mean for average consumers? In two words, probably zilch.
I'm not being overly pessimistic. There are indeed benefits to doubling the interface speed from today's SATA 3Gbps norm. For this, I'll turn to the tubes analogy: imagine you have a hard drive that delivers blazing read and write speeds. I'm talking about speeds in excess of SATA 3Gbps' current limit, roughly 300 MB/sec. In this case, the tiny interface tube is being flooded with more data than it can handle. Double the size of the tube, and you can tap into speeds between roughly 301 MB/sec and 600 MB/sec. Simple, eh?
The issue at hand is that the best of today's single hard drives--excluding their flash-based, solid-state cousins--simply cannot deliver this kind of performance, period. On top of that, I don't foresee a magnetic hard drive being able to even saturate the SATA 3Gbps spec anytime in the future. But for that, we have to consider two characteristics: a drive's burst speed and its sustained speeds.
Burst speed is a measurement of how fast information travels from the drive's internal buffer, or cache, to your system. While it's an important characteristic of a drive's performance, it only comes into play when the requested data falls within the cache--often less than 32MB of total information in today's top hard drives. More often than not, the drive has to seek out the information on the platters themselves. And the task of finding the information and transferring it to the buffer is measured by looking at the drive's sustained transfer rates. On the fastest of the consumer-grade hard drives, Western Digital's Velociraptor, the transfer rates don't even reach half the capacity of the SATA 3.0Gbps specification.
Seagate expects to have drives that use the SATA 6Gbps specification by the end of the year, but it's unclear what the company exactly means by this statement. It's been nearly half a year since Western Digital unleashed its Velociraptor drive and Seagate has yet to deliver a consumer-grade drive that tops its transfer speeds.
If Seagate is referring to the creation of a drive that's compatible with the new specification (which is itself backwards-compatible with SATA 3Gbps and SATA 1.5Gbps), I can get behind that point. But in no way will magnetic hard drives be ready to take full advantage of SATA 6Gbps speeds by the end of the year. And I'm not personally convinced, given the speed uptake over the past few years, that magnetic hard drives will be fast enough for SATA 6Gbps to matter within the next two to three years.
That said, it's an entirely different ballgame when you throw RAID and solid-state drives into the equation. String enough drives together--like twenty-four--and I can see a compelling argument for needing to increase the interface speed. And the use of RAID arrays in modern systems is more than you might think at first: For example, seven of PC World's top-ten Power PCs use RAID arrays for storage.
While the new SATA 6Gbps specification throws some additional features into the mix, like updated power efficiency and native command queuing, there's no word or benchmarks to show how this might influence the performance of modern hard drives.
If you're now holding off on a major PC upgrade because of SATA 6Gbps, well, you might be wasting your time. Solid-state drives will play a large role in filling the pipe of SATA 6Gbps, but it's difficult to predict the future point where their performance will outweigh their capacity and cost issues for an average consumer, especially when compared to magnetic storage. For your average, single-drive user, SATA 6Gbps just isn't that big of a deal.