One cable to rule them all: Light Peak

Device adoption will be interesting.   If the specs being leaked are right, then Apple's new Macbooks will have it's implimentation of Lightpeak but no USB3, only USB2.     What makes this interesting is that USB3 has a pretty wide head start, with USB3 drives, video adapters and SSDs already on the market.  Because USB3 and eSATA can share a port design, it's adoption in the PC laptop market has really come along.    Lightpeak is Intel's baby, and so with Intel and Apple, it will certainly have a big push behind it.   The question is, do we end up with another FireWire Vs. USB show down of sorts.

Originally there was some thought that LightPeak would be a single native interconnect, replacing SATA, Video, USB, Firewire, etc.  But that idea has went away.   Intel has come out and said that Lightpeak will not be the video standard, and instead has become big in supporting DisplayPort for that.   This is largely because getting another tech through a governing body for data protection is a bitch.

But as an interconnect, LightPeak has a lot of options.  It's 10Gb native data connection through standard Cat6 makes for a good way to change the way home networks, etc. work.  And it's hub and spoke method is very user friendly.

This is going to be very interesting because it's a great technology.   We'll have to see what vendors line up to get behind it.   Because while the paper technology is great, if not enough people end up supporting it (See: firewire) it never matures.

EDIT:  I've seen the stats floating around and they have to be wrong.   Someone has to have photoshopped this.  If it's an i5/2.13G, 320GB HDD then outside of Lightpeak there are no real upgrades here, in fact, using Intel HD3000 video is not using anything remotely "new", especially when the unit won't have an HDMI out, and Lightpeak, which replaces Displayport in the graphic won't do full DP1.2 compliance because it doesn't have enough bandwidth.  Between that and them listing an 8X DVDRW as the optical?   I realize Apple may still oppose bluray, but you can't put out a machine with the oldest optical technology still on the market, if you're going to stick with DVDRW, you know there are floating opticals at triple that performance in competitor laptops..

Give me an MBP with no optical drive whatsoever (I'll gladly take a higher capacity battery in the space) and I would be quite happy. Those specs listed above are attached to the 13-inch MBP, which still is using Core2 Duos instead if the i Series processors.

Quote:

Originally Posted by Michael_K_Sr 

Give me an MBP with no optical drive whatsoever (I'll gladly take a higher capacity battery in the space) and I would be quite happy. Those specs listed above are attached to the 13-inch MBP, which still is using Core2 Duos instead if the i Series processors.

Correct.   If I'm right this is a Sandy Bridge based laptop.  That will be interesting since Intel had to recall all of their 1st generation boards, so this is a good second bite at the apple (no pun intended).   The error in Sandy Bridge only impacted the older SATA ports and shouldn't be relevent here.

I'm still pondering this because I know the wife will want one.   But going from a Geforce 320M to an Intel HD3000 gains you nothing and loses you a step (about 30-40% of a step in newer games) in 3d, which may or may not matter.. it may get you some battery life, though, which might be the point.  I think there just aren't enough details.  The 15" one should be interesting.  

Quote:

Originally Posted by Sam Posten 

Soooo.... Light Peak premieres in new MacBook Pros Thursday and in ipad2 next weak?

GiggidyX2

http://news.cnet.com/8301-13924_3-20034900-64.html

I don't understand why you'd have Thunderbolt on the iPad.  Video out: Airplay (they just need to add Airplay input support to Mac OS X so it wouldn't require an AppleTV).  Move more of the synch to WiFi, but keep the wired option for large media files.  What else do you need that would require Thunderbolt?  You are never going to be hanging disc drives off the iPad.

The one thing that excites me is the possibility of long cable runs (optical version) -- however, these things seem to stay theoretical -- see: Display Port.

Daisy chaining is the number one reason we want Thunderbolt to strike USB dead. 

Quote:

Originally Posted by Ted Todorov 

Quote:

I don't understand why you'd have Thunderbolt on the iPad. 

Superfast syncing? Syncing my iPhone when I've added half a season of TV shows takes 15 - 30 minutes. Even a normal sync, with some photos and new podcasts takes a couple minutes to work through. Assuming its bandwidth limited and LP can drop it to a few seconds, life is nicer :)

Better industrial design? The 30-pin connector is pretty big. Replace that with a smaller, USB-style connector and you've more space to work with in designing the devices. It also allows for shrinking the device. The current Shuffle is limited by the connector, not the actual iPod hardware, I understand.

I would be shocked to see the iPod/iPad go to any sort of Litepeak exclusive connector, because you'd make it incompatible with every older mac and every PC.   LightPeak would require a SB or a 16 lane PCI-E connector, and there aren't any shipping on the PC side, and it wouldn't be an option for older macs.   So, while the idea of syncing faster sounds interesting, unless the iPad would have two docking ports, it won't happen.   and USB through a LightPeak adapter would still be limited by USB spec, which since it's a direct device would still be USB2, or just as fast as it is now.

Intel has now said (see Sam's other post) that it's a 3 Meter cable length.   So, not the length you may have been looking for.

Exciting tech, though.   I do want to say to those that think this will make transferring to an iPod or iPad faster, it won't.   The limiting factor there isn't the speed of the connector; USB2 can write as fast as the slower RAM based technology in them, and they have a USB2 host controller so it won't help there either, but even if they had upgraded their connector, you'd have to go to a more SSD ready format for data to notice the speed improvement.

Quote:

Originally Posted by Sam Posten 

This is dumb:

http://www.appleinsider.com/articles/11/02/24/intel_details_thunderbolt_as_exclusive_to_apple_until_2012.html

Yep.   This guarantees that it stays almost dead.   If you're developing periphials, do you develop them for something with an installed base that includes only brand new macs, or an installed base of every AMD PC in the last year and a half (USB3) or almost every PC and Mac since 2003 (USB2).  

I get it, Lightpeak (Thunderbolt) can be used as a breakout to go to all of those other things.   Which is cool.   But there is no inherent advantage in that at all if the only thing you have to talk to is yourself.

Intel now says there is NO exclusive agreement, they just don't have anyone else developing a board with it yet on the PC side, though they are free to do so.

http://www.engadget.com/2011/02/24/intel-refutes-apple-exclusivity-for-thunderbolt-i-o-lacie-and-p/

Quote:

Originally Posted by Sam Posten 

I also question that you have to buy a new machine to get it.  It wont be available for an option card that I can put in my Mac Pro?  Are you kidding me?

It could be possible, you'd need a x8 PCI-E card to make up for the bridge controller and the lanes.   But for most computers, that would be an X16 slot, which they are using for graphics.  So, it's hard to get that to work out.  Since a lot of the boards that have shipped, especially those that are H55, etc. aren't even PCI-E 8X in their 16X slots (for example, H55 boards are 4 lanes in the PCI-E X16 length slot) it would be no use at all.

It's a catch 22 for them.  

From your article:

You have GOT to be kidding me.   $100 to add 2 features that only a small set of units anywhere in the world can use?   How do you not learn from firewire?!?#   I will openly admit Thunderbolt is a standard with a lot of great potential.. but Firewire1 was so much better then USB1.1 it was stupid (400Mb vs. 10Mb)   So why did USB1.1 succeed?  COST.  

I hate to break this to Apple, but this isn't like an Ipod or an iPhone or something where people can look at it and admire the craftmanship.   Here, you're moving data for the most part at first.   For users, USB3 will seem as though it moves data "fast enough" along with eSATA which is basically free to anyone who already licenses SATA (which is everyone) at a near nothing cost... more then that, they are universally backwards compatible.

So you're going to introduce a new standard, and no matter how great it is, if the best you've got is 4 devices that have a starting cost of $1,000 and the they are data devices (4TB is $1K?  Are you SERIOUS?) 

What is the market for this?   The devices aren't big enough and they are single target (ie, it's not like they have multiple Thunderbolt connectors for multiple macs to access them) so they fail in comparison to say, Fiber connected SAN devices, while Faster then an eSATA device, the cost will be a killer for most...

Ugh.  They need some cheap, straight forward "MUST HAVE" Thunderbolt items that you can't get elsewhere.    Super-spendy attached storage isn't it.   Get a tumble scanner or a prepress to adopt Thunderbolt.   

Thunderbolt is a better standard then USB3.   There is no doubt about it.   But being a "better" standard of anything doesn't always matter if consumers don't have a reason to chose it. 

Sam, here's the link:

http://arstechnica.com/apple/news/2011/06/why-apples-2m-thunderbolt-cable-costs-a-whopping-50.ars

Sony is also coming out with Thunderbolt this summer; except it's rendition will use a proprietary cable they make for an outboard dock and GPU station only; so it will not be compatible with this station.

HP is also rolling out a thunderbolt port, it will also use a different connector type aimed at a dock.

In honesty, I can see this being used as an ultimate docking station solution.  In fact, if they just use it as an interconnect that a laptop could slip into for a full dock with video, and all other connectors it would be a great advantage compared to the docks that have risen up since PCs got faster.

But it also means Apple will be alone in their cable type for a while.  And at a $50 pop per cable (so per device) that's a big chunk of change, wiping out any low cost device for it.   In other words: don't expect to see a sub, say, $400 device that's Thunderbolt.   Because with a $100 cost for the chip and connector + fifty for a cable, it'd be a $250 device with that much in add on cost, so it wouldn't be very tempting.

This problem though isn't Apple's, it's Intels.  Intel could have pushed Thunderbolt out as optional components on new Intel branded motherboards, or options for new Sandy Bridge options that would have put it in the OEM market.. broader adoption would have meant lower prices and more devices.   They didn't.  And now the cheese stands alone with the super cool toy.  And that's Intel's fault.