IEEE signs off on 400 Gb/s Ethernet development As we foreshadowed last month, the The Institute of Electrical and Electronics Engineers (IEEE) has green-lighted development of 400 Gb/s Ethernet. Now known as IEEE P802.3bs, the nascent standard now has an official IEEE task force working to make it a reality and a goal, stated in this IEEE document (PDF) , to “Define …
You realize the reason that gigabit ethernet is so affordable is because it is useful/needed in the consumer market, and 10Gbps is not.
Just because you want something to come down in price shouldn't mean that consumers should start having to spend more for stuff to give it 10Gbps ethernet ports they have no use for.
That would be a poor argument. When we went to 9600 the benefits were huge and obvious. It was definitely needed, not least of all because it reduced the cost of batch-data phone calls by so much....
Consumer demand drives vendors. Not the other way around. Exception being anything with an apple on it.
Actually 10Gbps Ethernet could have its place in the consumer market. It could be used to transfer high definition video at low latency. For example you could have a single computer in your basement/closet which powers all the screens in your flat connected via fast Ethernet and small dumb terminals which may consist of nothing more than an ethernet controller and a framebuffer.
@Christian Berger, Fast Ethernet is slower than gigabit Ethernet, I think you probably meant connected via 10GbE?
Either way, video is predominantly going to be coming over the Internet in the coming years so 1GbE is more than adequate internally for everyone other than those same bunch who insist on using CD or FLAC while all the normals go and use MP3 and streaming services. I expect I'll be down voted for saying that but lets face it, society usually votes for easy and cheap over better and pricier.
Sure, if you want to sling around uncompressed HD video. That's what, 0.2% of the consumer population?
It is easy to come up with a list of reasons why consumers should need faster networking, faster CPUs, etc. but not so easy to come up with mass market reasons. Find a killer app that needs something faster than 1Gbps networking, and cheap 10GBE interfaces will follow.
"Any word on zero byte latency? This is a number that does not seem much downward movement."
As it should be. The laws of Physics are rather unbending. It still takes at least 3.25 nanoseconds to traverse each metre of interconnect length no matter how you transmit the data.
Maybe when Ethernet is transmitted via quantum teleporting we might be able to look forward to less latency. I doubt it, though.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
