Now that we’ve tolled the death knell for open, expandable computer systems in Apple’s line up, it’s time to put forth a series of discussions about what we at Meta Media see as the forthcoming technologies that will be the critical components in future collaborative workflow infrastructures.
Firstly, it is our belief that Light Peak will be the singular peripheral protocol on all future Macs. It’s been widely reported that this technology’s debut is months away. Further, it’s been reported with less certainty that Apple brought the technology to Intel in order to develop it and usher it through the standardization process that is critical to industry adoption. Apple will then most probably insist that it have exclusive rights to release computers carrying Light Peak first, only allowing Intel to release it on its reference PC motherboard after these machines have come to market.
Like USB before it, Light Peak will be primarily a local bus protocol for communication with devices external to the computer, but it improves upon USB in several important ways.
First, it will have both copper and optical connectivity specifications, allowing long distance connectivity (100m is being discussed) as well as high speed data transfer. It’s first specification will allow 10Gbps with a road map to push that to 100Gbps and beyond.
Second, it will ensure that communication with any kind of peripheral device be maintained at the speed necessary for that device without affecting the communication speed of other devices. You’ve probably been the victim of the most serious deficiency in the design of USB: your connected hard drive copies files at a significant clip until you plug a mouse into the same bus and the speed plummets. Light Peak will reportedly do away with this issue by supporting several levels of simultaneous communication speed, eliminating the need for all devices to slow down to the speed of the simplest device on the bus.
These two simple innovations will inspire peripheral manufacturers to adopt it rapidly, although it will take their R&D departments a little while to come up with the designs. This means that peripheral availability will be slightly delayed after Light Peak’s release, with only a handful of eager manufacturers at the starting gate. The first players will be the ones with close relationships to Apple and Intel (or both), who will endure the draconian procedures that Apple uses to keep its final designs secret. Then, a slow trickle of others will follow as sales reports come back and competitors enter in when the water is warmer.
A delightful side effect to this will be the realization that with Light Peak’s core features, there will really be no need for other peripheral protocols such as USB, SATA, eSATA, SAS and Firewire. It could also be argued that PCIe will not be too terribly missed, as long as we get speedy delivery of devices that replicate the functionality of PCIe cards we used to put inside the machines.
Imagine a device similar to the AJA IoHD that can handle up to 2K and perhaps 4K uncompressed images at appropriate streaming frame rates, all available to a desktop or feisty laptop with a simple connector. Firewire ports will be replaced with Light Peak, enabling the transfer of uncompressed data in addition to compressed streams.
And, similar to the adorable USB to Ethernet adapters that are currently available for the MacBook Air, we may very well see Light Peak adapters that allow the adoption of communication protocols such as 10Gb Ethernet.
One thing is certain in our view: the box that Apple makes, into which all of these magical devices will connect, will be closed with tight seals. What Light Peak affords Apple is to move forward with their vision of walled garden systems across their entire hardware product line without shutting out any particular user or market segment.
You might have thought at this point, “what about USB 3?” It will be a competing protocol for certain. It will have legs in the market because it will be easier for manufacturers to adopt, since it is essentially the same protocol as USB 2 with faster speeds and no further innovation. In our opinion, Apple will forgo its adoption (in fact, Apple could have embraced it during several recent product releases and has clearly not done so yet). The issue here is that, as with many standards wars that bubble up in this industry, some peripheral manufacturers will be torn as to which bus to adopt. USB 3 will most likely come to be known as the “poor man’s Light Peak,” being that we will most likely see it on less expensive hardware such as netbooks and entry level PCs. Those manufacturers that will continue to feed off of the Apple ecosystem, however, will have to spend the time and money to make their devices Light Peak compatible.
Lastly, if the PCIe slots are really going away, what of graphics cards? They too will be intertwined into the motherboard, as they are on all the laptops, iMacs and minis (and even the Xserve!) today. And even though Light Peak boasts the ability to communicate with digital displays, most Display Port specifications require speeds beyond 10Gbps. Therefore, we see the Display Port continuing to drive the displays of future Macs, independent of the Light Peak bus.
As for networking ports, we’ll tackle that in the next article...
Nice Article! Maybe you'll get to this with networking in the next installment, but, I was wondering if light peak connections can be assigned an IP address similar to today's firewire. Will be seeing sans with light peak connectivity in 3 years?
Posted by: Keith Koby | November 15, 2010 at 08:11 AM
Good question! Light Peak at present is a bus just like USB. Therefore, it will not be addressable nor will it be able to link computer to computer, just computer to peripherals. As I mentioned, we might see Light Peak adaptors that allow connectivity to networking systems and perhaps SANs
Posted by: Matt Geller | November 15, 2010 at 09:22 AM
The mac pro motherboard does not have the room for good video on the motherboard. You need a card to get good video maybe in a laptop like MXM slot.
Posted by: Joe12345 | November 26, 2010 at 05:20 PM
I think when Apple wants the finest offering of ATI, directly on a non-expandable "Pro" Mac, it will make the room.
Posted by: Matt Geller | November 27, 2010 at 06:26 AM