Updates to this story
While the world and its dog has been running stories about Chipzilla's new Thunderbolt technology, AMD says that it is not particularly impressed.
While we would not expect AMD to unwrap the bunting for its rival's latest efforts, it is doing its best to dismiss the Thunderbolt input/output technology as just another proprietary standard with not a lot of chances to be adopted widely.
A spokesman for AMD told Xbit that AMD could not see how Thunderbolt brings any tangible improvements.
At issue are the existing standards which seem to be better than what Chipzilla is offering. Thunderbolt can manage 10 Gb/s bandwidth if the wind is behind it and it is going downhill.
This means that it does not substantially outperform current generation I/O technologies and sometimes even offers lower bandwidth.
Existing standards allow good connectivity as a bonus and can still go faster than any of the peripherals that are available.
DisplayPort1.2 standard offers up to 17Gb/s of peak bandwidth for displays. The total bandwidth for a Thunderbolt channel is only 20 percent higher than one PCI Express 3.0 lane and about 52 percent higher than a single USB 3.0 port.
AMD said it was daft to use the mini DisplayPort connector for the technology as this reduces bandwidth available for displays connected to the mDP port.
AMD said that using Thunderbolt in the DisplayPort connectors decreases the bandwidth available for DisplayPort, reducing the bandwidth available for various multi-display configurations.
While Seagate Technologies and Western Digital do support Thunderbolt and plan to introduce HDDs featuring the interconnector later this year, only high-speed solid-state drives will be able to actually use the higher bandwidth.
Probably.
Maybe I am loosing it but can you tell me what the difference is between this sentence
"AMD said it was daft to use the mini DisplayPort connector for the technology as this reduces bandwidth available for displays connected to the mDP port."
and this sentence
"AMD said that using Thunderbolt in the DisplayPort connectors decreases the bandwidth available for DisplayPort, reducing the bandwidth available for various multi-display configurations." ?
@ time2go
I can see why AMD is marginalizing this tech from Intel (because they don't have it) but anyone else who looks at what "Light Peak" (the Intel codename for Thunderbolt - they should have stuck with the codename in my book) can't help but be impressed.
I agree the first electrical iteration of Light Peak is not as exciting as getting a fiber optic connection directly from the chipset without the additional latency of a separate bus chipset to convert the electrical signals into fiber optic signals. However, the optical version of Light Peak is the future in peripheral connection. I can't see how an electrical connection can try to match the latency and speed of a fiber optic connection directly from the peripheral to the chipset. I would guess the connection will eventually be straight to the CPU. Your comments remind me of how 15 years ago some refused to believe serial communication would ever be better than parallel port communication. Where are parallel ports and SCSI connectors now?
The reason this technology is so cool (and AMD has sour grapes) is because it is being done in Si which is an indirect bandgap semiconductor (meaning it is very difficult to get light to absorb or emit). Optical communication with a semiconductor is normally achieved using a direct bandgap material like GaAs. GaAs is incompatible with Si. A lot of very smart people have tried to solve this fundamental problem in Si for years. The fact that Intel now has a solution is going to dramatically increase the adoption of fiber optic communication and we will all be better for it.
Try not to assume that just because Intel has a technology you don’t understand that the adoption by OEM’s means they were coerced in to it. The OEM’s may know more than you. Would you prefer Intel stopped generating these innovations just so that they will be less of a competitive threat? Where would Moore’s Law (and the computer you are using) be today without Intel?
Errr .. I beg to differ. I hardly had to try at all ... I'm not impressed.
It looks like a technology designed to solve a problem that was already solved.
If this had come out when we only had USB1 or maybe even USB2, then Intel would have had a red hot winner.
However USB3 has been out for a year and meets the need of users already in this 'high speed' transfer area.
@time2go Agreed. I have to say I am always wary of new Intel technology that has been developed unilaterally, and then is not immediately opened up for general use.
They have a habit of trying to use 'add on' proprietary technologies to put apply pressure to corporate buyers.
I say to Intel ... if you believe it is better than USB3, then show your faith by proposing it in a similar fashion. USB3 is vendor neutral and available to all implementers without restriction. Will Intel do the same? or again just try to Muscle the market.
Most computer users were rarely hitting the limits of USB2. Those involved in regular large media transfer (ahem cough) might like the sound of it, however there is a kicker, Intel would certainly incorporate DRM / HDCP restrictions (it has a business subdivision that sells it remember) if it could grab over half the market.
Personally I think there is little danger of USB3 being displaced for reasons mentioned. It'll probably just give a few MacBook Pro users something to chirp about other than Big Brother or Celebrity gossip.
Yes, they 'have' the market now... If you own an Apple, anyway. :P