Light peak, better than USB3.0

USB 3.0, Light Peak, and Upgrading


  • Total voters
    27
usb 3.0 will be around for a while.. probably in the next 2 years it will become the standard instead of usb 2.0 ports.. so either way with light peak out or not usb 3.0 will still be around..
 
agreed. USB has been around far too long and there are too many devices out that it isnt going to go anywhere anytime soon.
 
usb 3.0 will be around for a while.. probably in the next 2 years it will become the standard instead of usb 2.0 ports.. so either way with light peak out or not usb 3.0 will still be around..

Agreed.

But I think light peak will change the way we use computers much more than usb 3 will.
 
Light Peak will not best USB 3.

Light Peak will threaten Gigabit Ethernet, eSATA, and the myriad of display connection standards we have to deal with.

Light Peak does not offer a serious improvement over USB 3. The bandwidth demonstrated is 10 Gbit, and although higher speeds are possible, the fact remains that higher-speed fiber is very expensive (as is the signaling hardware required to drive it at those speeds). That bandwidth is not enough to make people jump-ship (just look at how far Firewire 800 got versus USB 2). Also, Light Peak has no way to deliver power over the optical cable alone, which makes USB 3 very attractive for external drives (4.5w max can power SSDs and 2.5" hard drives). And of course, USB 3 has a backwards-compatible connector, which will make people VERY happy.

Light Peak is also trailing in adoption. There are ZERO Light Peak products on the market, and the market is side-stepping Intel's non-support of USB 3 by adding it motherboards themselves. AMD will be releasing native support in SB850 this year. Intel is not planning on revising chipsets until 2011 (isn't Intel's chipset monopoly great?) be it for USB 3 or Light Peak, so I don't expect any market push for optical tech until next year.

Light Peak will be a nice improvement over Gig Ethernet. The two things holding 10 Gig Ethernet back from the market are 1: the length of the drop capable of 10Gig speeds over existing cables, and 2: the amount of power required to drive the physical interface. The power has grown with each generation, because a higher transfer rate over the same copper requires more power to increase the signal to noise ratio (1 Gig Ethernet uses aboout 2-3w, which is a lot for a mobile device, and 10GBase-T is at 5w right now). Optical fixes both of these problems, with much less than 1 watt required for the Light Peak 10G physical layer. Light Peak also has an upgrade path, whereas 10GBase-T is pretty much the end of the road for Ethernet over copper.

As for the poll: if you are building mid to high-end, don't wait to upgrade to USB 3 until support is integrated. Intel obviously doesn't care about USB 3, or they would be introducing a chipset with support this year. Mid-to-low builds will get support with AMD's SB850 later this year, so then you'll have support across the range.
 
Last edited:
I'm not sure whether there is an option besides "Upgrade to USB 3.0 when its officially integrated by Intel/AMD"... For (a), you need USB3.0 stuff, which we don't have yet, and (c) why would you "wait" for a long-distance-time-frame, peripheral interconnect bus?
 
Any motherboard you buy in the next few years will have USB3 whether you like it or not, and it may or may not have LightPeak, depending on how that turns out. Even if LightPeak is undeniably awesome, USB3 will not go away quickly because of all the legacy components.
 
Optical cables can be a real pain. I'm not in a big hurry to replace a bunch of durable USB and Ethernet cables with fragile fiber...

I actually like technologies like ESATA, because it allows me to have an external drive controlled by the same internal SATA controller that controls the rest of my drives.
 
I wouldn't look at USB 3.0 as merely a speed upgrade. There are many changes that don't hinge on the ability for the speeds to reach it's potential.

-You charge your mobile devices faster,
-Duplexing so you're thumb drive doesn't take a shit on you when you transfer songs while listening to it
-Duplexing will increase your keyboard's ability for feedback whether it be a mini-lcd or force feed-back, or perhaps even a mini-oled touchscreen.
-Even a minor speed improvement crosses the threshold for 1080p streaming
-I'm guessing here but increased power and a portable ssd gets you a single usb 3.0 connection instead of a two wire solution
-3.0 supports power management, reducing energy usage

I think it's a solid investment, regardless if Intel/Amd have support right now. USB or LP add to the user experience, they don't define it. I wouldn't get puckered over the decision, I would just try to take advantage of as many upgrades as i can afford when making a purchasing decision.
 
I remember people saying Firewire would kill USB when it first appeared on motherboards.

About 4 motherboards later and I've never used the Firewire port on any of them and I still use every single USB port there is.

To kill USB it would have to be faster, cheaper, and backwards compatible with the billions of USB devices people are already using.
 
I voted upgrade now. I actually have an Asus add in card I haven't installed yet, but I plan to use USB 3.0 until something else comes along that changes my mind. Until it's as widespread and inexpensive as USB, I'll stick with USB 3.0.
 
One thing that many people don't seem to think about with completely new technology implementations such as LightPeak is the cost to implement and manufacture. Like Limahl said, it would have to be faster, cheaper, and backwards compatible with the billions of USB devices out there. LightPeak may take over for such items as storage devices but I would see it more in an enterprise item where you can be running a 10GigE network within a building and want the fastest possible connection for small items like NAS, Tape Backup Drives, etc.
 
Back
Top