Light Peak: Cost, copper, and crippled

[red1776]

Light Peak (Thunderbolt) is not being met with open arms by the likes of Apple and the worlds largest PC maker HP. It seems that Thunderbolt is on the back burner for now.
From ExtremeTech.com

http://www.extremetech.com/article2/0,2845,2385615,00.asp

 

[dividebyzero]

Article sounds fairly poorly thought out. Reads like an introduction to a more in depth analysis...but then I turned the page and...
For an interface thats "on the back burner for now" it will be shipping on every new iMac and every MacBook Pro
So, the two obvious questions that come to mind are:
1. Will Apple owners be put off by the extra cost of a Thunderbolt enabled peripheral ?, and...
2. Should PC owners be sad that Apple users are bearing the brunt of high introductory prices, R&D costs, debugging (if necessary) and limited availability ?

 

[red1776]

Article sounds fairly poorly thought out. Reads like an introduction to a more in depth analysis...but then I turned the page and...
For an interface thats "on the back burner for now" it will be shipping on every new iMac and every MacBook Pro
So, the two obvious questions that come to mind are:
1. Will Apple owners be put off by the extra cost of a Thunderbolt enabled peripheral ?, and...
2. Should PC owners be sad that Apple users are bearing the brunt of high introductory prices, R&D costs, debugging (if necessary) and limited availability ?

I was wondering if the initial offering as copper was going to kill light peak by perception. I hardly think it's "crippled" as this dude seems to think, but HP turning up their noses isn't going to help adoption.
I now wonder how long it will be set back by USB 3.0.?

 

[dividebyzero]

I was wondering if the initial offering as copper was going to kill light peak by perception.

As has been ably demonstated on these forums amongst others, most people don't have a clue about the technology and will always defer to what they are comfortable with. There is also going to be a perception amongst a percentage of people that Intel=Bad. You could then answer that both HDMI and DisplayPort count Intel amongst their developers. The same group are also going to tout DP's "royalty free" status (as is Thunderbolt) while being blissfully unaware that the DP specification doesn't rule out compensation to contributors or IP holders in future
At the moment Apple is the company offering Thunderbolt, much as they were a prime mover behind DP's adoption. Percentage of Apple users who likely know anything about Light Peak spec ?
My guess is that Light Peak/Thunderbolt will meet resistance in the consumer space so long as it remains expensive (see DisplayPort) and esoteric (Firewire), and enough ignorant people are willing to shell out for every incremental advance in HDMI and DP

I hardly think it's "crippled" as this dude seems to think, but HP turning up their noses isn't going to help adoption.

Probably not, although how the other OEM's respond will dictate how quickly the uptake is I would think. HP are probably anti-Intel-anything I would think after hitching their server business to the Itanium (death)star. With Apple onboard the tech gets a platform in the consumer space, IF a large OEM comes on board- say Dell or instance (who pretty much pioneered consumer 2560x1600 and DP) who might-for example- decide to market a WQUXGA, 4K or HXGA /T-bolt panel, then you'd think that HP would rethink their position. Uptake would be slow in any case-it always is, until it reaches a point where it becomes a standard feature.

I now wonder how long it will be set back by USB 3.0.?

Or USB4.0, or HDMI 1.5, or DisplayPort 1.3 etc, etc.
At a guess I'd say the moment someone produces a saleable 4K+ monitor and then realizes that the DP cable needs to less than 30cm long to run it. Remember that DP is limited to 2m for 2560x1600, and 3m for 1920x1080 (How's that Holodeck lookin'?).
I could well see Intel bringing out a Light Peak internal optical interconnect turning I/O latency into ancient history. No doubt TS's forums in 2015 will be full of "OMG!!! Y dint any1 think of dis sooner LOL"

 

[red1776]

Or USB4.0, or HDMI 1.5, or DisplayPort 1.3 etc, etc.

I was hoping it wasn't that far away...I mean, I don't make use of USB 3.0 found on my Asus C4F yet, I just hate new tech being sidelined by false perception.

At a guess I'd say the moment someone produces a saleable 4K+ monitor and then realizes that the DP cable needs to less than 30cm long to run it. Remember that DP is limited to 2m for 2560x1600, and 3m for 1920x1080 (How's that Holodeck lookin'?).

I need new cables......

I could well see Intel bringing out a Light Peak internal optical interconnect turning I/O latency into ancient history. No doubt TS's forums in 2015 will be full of "OMG!!! Y dint any1 think of dis sooner LOL"

...but apparently not for another three years.

 

[dividebyzero]

Yeah, great how the prospect of something that might eventuate a few years in the future becomes some kind of balm.
Think AMD wish they'd have taken the long view regarding handheld and ultraportable markets? pro graphics and server support ?, IBM and Co. gate last over gate first ?
There's a reason that some companies have a virtual monopoly in certain markets, and I wouldn't attribute the lions share to dumb luck

 

[Archean]

The idea in its original form (ironically proposed by Apple) i.e. one socket/interface for all connectivity is very commendable one, but sadly it has been ditched with the advent of thunderbolt’s new socket interface. I think the question should be asked is that when lightpeak originally had socket compatibility with UBS 2/3 etc. then why do they need to come up with a new socket design/implementation?

Note: I am rather focusing on the technical reasoning behind this decision.

 

[dividebyzero]

It was Intel's (and Apple) intention to use the USB2.0/3.0 socket. The USB-IF torpedoed the plan:
"USB connectors are not general purpose connectors and are not designed to be used in support of other technology applications or standards or as combo connectors."

I think this article is a better reasoned piece (IMO)

It is certainly prophetic.
How ofter does a consumer computer enviroment make the News section of a mainstream news site, rather than the Tech pages? I'd think that an article like this is going to more influential than most of the tech sites forums combined. I'm kind of doubting that Hewlett-Packard's builds would warrant the same mainstream attention.

 

[Archean]

Thanks for the link DBZ.

So in essence, a superior solution is being stopped in its tracks by other factors. But I will still side with Apple/Intel in this, a user will benefit with a single interface for all sort of connectivity IMO. So if they are 'forced' to come up with new interface/socket, I guess Intel will need to twist some arms to have their way (in the interest of everyone).

 

[dividebyzero]

If USB-IF want to preserve the USB-for-USB there isn't a whole lot of options . If VESA (DisplayPort) are happy enough to open their socket to Thunderbolt, then there is some broad continuity, if not, and a new socket is needed, I wouldn't see having one more being a major hindrance- look at the rear panel of most motherboards and you'll see eSATA, USB 2, USB3.0, Firewire (400 and 800 in some cases), TOSLINK, Coaxial, GbLAN, and possibly three of HDMI, DVI, DP and D-Sub15.
In fact having the extra socket or two might add to it's appeal. There's a certain element of PC ownership who seem to view Number of cables exiting a box = higher grade of enthusiasm. Hence (I guess) the mandatory cable porn shots in high performance computing puff pieces.

Not sure about the "arm-twisting" that Intel might need to do. Intel were instrumental in bringing most of the connectivity most of us now take for granted. I don't see Thunderbolt be a whole lot different from, say, S-ATA over P-ATA. The only difference now is that some of Intel's partners want their particular interface to be the default standard, and probably see their possible future roles as marginalized if Intel keep Thunderbolt to themselves. If the optical connect was solely for peripheral connectivity I could see the point to the argument, but as seems painfully obvious, Intel are in all likelyhood going to integrate optical interconnects into (at least) server based systems it would seem fairly prudent to keep that IP and patents to themselves. As far as I'm aware, Intel does not have a lock on optical digital data transmission- but they seem to be the only company willing to invest R&D manpower and funding. Why give away your assets?