Since when has 6Gb/sec = 600 Mb/sec?

[Guest]

Ive seen a lot of hard disks advertised as being 6Gb/sec hard disks and yet when I look at the spec it says they are 600 Mb/sec - surely this cannot be right?

For example:

http://www.amazon.co.uk/Seagate-ST2000DL003-Barracuda-3-5-inch-5900RPM/dp/B004C0WVP6/ref=sr_1_3?s=computers&ie=UTF8&qid=1310374385&sr=1-3

Anyone?

 

[tobensg]

Hi. The reference (Gbps) is Gigabits/second, not Gigabytes/second (GB/s). 6 Gigabits maximum throughput on paper is 750MB/s, in reality its 600MB/s.

 

[tobensg]

MB/GB/gb Calculator link;
http://web.forret.com/tools/bandwidth.asp?speed=6&unit=Gbps

 

[Guest]

Thanks for the replies I now know the difference between Gb and GB so I've learned something new today.

 

[tobensg]

Your welcome

 

[WyomingKnott]

One point I'd like to make: Although a byte has 8 bits, the bus uses encoding that expands each byte to 10 bits. This helps overcome noise and balance DC so that there is no net voltage between the data pins.

So 10 bits/sec of interface speed gives you 1 byte/sec of data speed.

Some interfaces are using coding with less overhead than 8/10 these days. I forgot the context in which I read about this; was it gigabit Ethernet?

 

[tecmo34]

One additional thing to note... It comes no where near the 600 Mb/sec | 6Gb/sec throughput of the SATA III controller. It doesn't even saturate the SATA II controller.

http://www.storagereview.com/seagate_barracuda_green_2tb_review_st2000dl003

 

[Guest]

Very interesting if I read the charts correctly the maximum this supposedly 6Gb/s drive can practically read/write is 1.5Gb/s so you could easily use it on a normal SATA 3Gb/s port no problem. Makes you wonder how these manufacturers can get away with rating it their drives as 6Gb/s when they cant actually achieve that speed in real terms.? Sounds like misrepresentation to me.

 

[sminlal]

Some interfaces are using coding with less overhead than 8/10 these days. I forgot the context in which I read about this; was it gigabit Ethernet?

Engineers who design data transmission protocols at the physical level generally quote raw bit/second rates at that level and ignore everything else. But if you go up the protocol stack you find that the actual packets (be they Ethernet, SATA, USB, etc. etc.) contain non-trivial amounts of metadata such as command type, size, checksums, etc. And each additional layer adds another level of overhead (especially true in networks because of the deep protocol stack). So the real-world transfer rates for actual application data never match what you'd extrapolate from the transmission speed alone.

 

[sminlal]

Makes you wonder how these manufacturers can get away with rating it their drives as 6Gb/s when they cant actually achieve that speed in real terms.? Sounds like misrepresentation to me.

It's perfectly fair for them to claim that the drive communicates at 6Gbit/sec because that's how fast the "wire" actually runs. But it is very misleading for them show it in huge letters on their ads and packaging IMHO. It's kind of like an auto manufacturer trying to push "190km/H" as the most important feature of a car. Everyone's familiar with cars and knows that you'll probably never actually achieve that speed - but most computer users have a much lower level of technical understanding and the manufacturers are trying to take advantage of it.