Do we
really need to keep revisiting the need for 10 Gig?
This. Happens. Every. Single. Time.
HideOut :
So I've ran gigabit in my home for about 15 years.
Maybe you haven't noticed this, but a lot's changed in the past 15 years! About 5 years ago, SSDs capable of over 500 MB/sec became fairly commonplace. More recently, NVMe has enabled SSDs to reach blistering speeds of over 2 GB/sec!
Not to be left out, ever-denser HDDs' media transfer rates stretched to 200 MB/sec, and beyond.
HideOut :
Even today with streaming 4K it don't even come close to using the full bandwidth of gigabit. Maybe if several people were pulling off of a NAS at the same time, sure.
Yeah, so I'd agree that streaming 4k video is a bad example. However,
editing 4k video hosted on a network drive, or
copying 4k video files are
fine examples of reasons you'd want something faster than gigabit.
HideOut :
But individual appliances dont even come close.
Maybe not your typical 10-year-old NAS, but it's not even hard to hit the 1 Gbps ceiling with more recent models. Why do you think the higher-end units started featuring 10 Gig ports, even a few years back? If costs were lower and the switches more commonplace, I guarantee even more NAS' would have them.
About 6 years ago, I built a home fileserver with an AMD Phenom II x2 (dual core) that can sustain about 325 MB/sec reads and 275 MB/sec writes, IIRC. That was with 5x 1 TB HDDs in a software RAID-6, yielding 3 TB of usable capacity. Assuming the recording technology and rotational speed is held constant, media transfer rates should increase proportional to the sqrt() of the platter density.
In summary, a
single user, copying
single files between SSD-equipped computers can
easily benefit from > 1 Gigabit. That's probably the most common use case. The need is probably more compelling for users of NAS and fileservers, since they're copying much more data, even if the actual storage is a bit slower. The world is
very much ready to go beyond the limits of gigabit Ethernet!