AMD's Lightning Bolt: Low Cost Thunderbolt Alternative for USB 3.0/DP
by Anand Lal Shimpi on January 12, 2012 9:11 PM EST- Posted in
- Trade Shows
- CPUs
- AMD
- CES
- Thunderbolt
- CES 2012
- Lightning Bolt
One of the things I love the most about AMD is the balance it provides to Intel. While I've spent much of CES looking for Thunderbolt products and lamenting the cost of controllers and devices, AMD put together a concept it calls Lightning Bolt.
Lightning Bolt is an AMD technology that can deliver USB 3.0, DisplayPort and Power over a single cable with mini DisplayPort connectors. I saw the technology demonstrated live, however I wasn't allowed to take any photos.
The technology is designed to be very simple and affordable. On the notebook side is a mux that combines power, DisplayPort and USB 3.0 into a single DP-like cable. The other end of the cable would connect to a Lightning Bolt breakout box that would provide USB 3.0, DisplayPort and power ports.
The cable is a standard mini-DP cable with changes on only two of the pins. AMD's goal is to enable affordable, single-cable docking stations for notebooks. The cost of the mux and associated components on the notebook side would be minimal, around a dollar. The mux would eventually be integrated into a notebook (the AMD demo had them external for demo purposes) and all you'd see is a mini-DP interface with some sort of indication that it was a Lightning Bolt interface. Given that it's a simple mux on the notebook side I'd assume that it would be possible to enable miniDP passthrough and display Lightning Bolt entirely if you wanted to.
There are performance and power limitations to this design. AMD claimed USB 3.0 transfers would be faster than USB 2.0, but not full speed. No word on how much power you'd be able to send over the interface either. As far as the docking stations go, AMD expects that they'll cost about as much as a USB 3.0 hub.
Lightning Bolt won't be ready in time for Trinity's launch in the middle of the year, but AMD hopes to have it on the market by the end of the year.
54 Comments
View All Comments
cdillon - Tuesday, January 17, 2012 - link
Physics fail. Signals travel down copper at somewhere between about 65% the speed of light for coaxial cables and 97% for a free conductor. Not sure were you got a 100X difference from.The "point of fiber" is absolutely bandwidth. The maximum theoretical bandwidth limit is far, far higher for fiber optic cable than a copper wire (again, those pesky physics).
Easily isolated fiber optic cables also do not induce nor pick up interference from the surrounding environment, while every copper wire doubles as an antenna. Noise immunity is a giant hurdle in any copper communication link. Fiber links simply don't have to deal with it.
mino - Tuesday, January 17, 2012 - link
Wrong, problem is frequency.You can make cheap copper cable but you need complex demuxes on the ends (which NEED time to do its work) plus you have to run at low freq.
TB would have been wonderfull had the actualy been a primarily fiber interface.
You take optics from TB and you get an inflexible useless eSATA/ePCI/eSCSI ... you name it.
A shame.
mino - Tuesday, January 17, 2012 - link
Then, he is actually advocating FOR LightningBolt(whatever they anme it) as Thunderbolt is an overkill for 99% use cases./me would love commodity TB but, well with intel holding the patents none is gonna mass produce it (on the device side), zero royalty ODM's a**.
mino - Tuesday, January 17, 2012 - link
USB3 has as much in common with USB1/2 as PCI has with PCI express ...