Next-Gen Thunderbolt: Twice as Fast, But a Different Connector 178
Details have leaked about the next iteration of Intel's Thunderbolt connector. The good news: bandwidth will double, going up to about 40Gbps from its current 20. Power usage will drop by half, and it'll support PCI-e 3.0. The bad news: it uses a redesigned connector, and will rely on adapters for backward compatibility. From the article:
"Doubling the available bandwidth will enable next-generation Thunderbolt controllers to drive two 4K displays simultaneously, where current controllers can only drive one. The new controllers will allegedly be compatible with a variety of other protocols as well, including DisplayPort 1.2, USB 3.0, and HDMI 2.0. Intel will offer two different versions of the controller—a version that uses four PCI Express lanes to drive two Thunderbolt ports and an "LP" (presumably "Low Power") version that uses two PCI Express lanes to drive one port."
In other news (Score:5, Interesting)
The 10 people affected by this bus imrovement went out to celebrate but were hit by a car going twice the speed limit.. Oh the humanity!
Seriously though, I like to consider my needs a non-professional leading on the bleeding edge (2x 2560x1440's) But I don't even own a thunderbolt port, and unless some amazing peripherals come along to change my use case, I don't see that changing soon.
All I want is:
1. standard bus standard which can drive anything
2. said connector/cabling comes in 3 sizes from really really tiny cell phone variety to honking large clicking in connector that can't break
3. That is future expandible to whatever for the next 10 years minimum
4. No IP which prevents competition in said space except for standards bodies who's potfolios are both fair and unbiased in licensing terms
Addendum I. Monster cables is specifically banned from ever producing said cables for ever
Nice to have's
5. Fibre option
6. Broadcast based networking support
7. Bus QOS control
8. Standard descriptive naming (NO BS marketing names like super-speed, hyper-active speed, high definition bandwidth, etc. )
9. Support host wake/power-on
10. Support at least bi-directional communications so I can plug in Bluetooth/IR/Wifi/etc.. message receivers and have if not chipset, at least OS support for pluggable and routable support for input methods without BS proprietary support all over the place
Re:New connector great thanks (Score:4, Interesting)
What really made USB popular was when USB 2.0 hit, and again Apple had nothing to do with it. PC manufacturers started adding more USB ports and it became easier to attach a number of devices. Who would attach a keyboard and mouse when their machine only had two USB ports? HP in particular was one of the companies that helped develop USB 2.0, and they started moving to USB keyboards and mice as standard around then too.
What really made USB popular was that it was cheap shit. There was practically no licensing cost on the interface, and all you needed was a few pennies worth of host-driven IO. When it makes no difference to your bottom line, of course you put a dozen of the fuckers on a machine. Your mouse and keyboard really need no significant performance, so who cares if it's on a slow, CPU-hungry interface, and even things like (consumer grade) scanners and printers benefit more from a small, hot-attachable plug than they do from a hardware driven interface.
When you start trying to use it for storage, or cameras, or network interfaces, or monitors, or anything with any real data rate, now the cheap shittiness of the interface of the interface rears its ugly head, but it's too late, because it's already the pervasive peripheral interconnect. This is why Apple was trying to hold out with Firewire over USB