Yesterday Intel announced its updated Redwood Ridge Thunderbolt controllers (adding 4K/DP1.2 support, lower power operation and slightly lower BOM cost) as well as next year's Falcon Ridge family of Thunderbolt controllers. I assumed that Falcon Ridge would ship alongside Broadwell in late 2014 but I just got word that the new Thunderbolt controllers will begin production by the end of this year and ramp volume through 2014. I don't have more specific dates than that, but it's still good news.

All Intel is saying about Falcon Ridge is the spec includes a speed bump to 20Gbps per lane up from the current controllers with 10Gbps per lane. Intel's own presentation refers to a single channel/two lane Falcon Ridge controller, delivering no additional bandwidth above existing TB implementations (although there may be practical advantages since you can't easily aggregate all TB channels today). I would assume that we'd see even higher bandwidth implementations but Intel isn't talking about any of that today.

POST A COMMENT

31 Comments

View All Comments

  • epobirs - Tuesday, April 16, 2013 - link

    Thunderbolt IS an external PCI-e connection. It is PCI-e over a cable and uses the same protocol. This is why it is so easily applied to things like external GPUs. Splitting off the DP signals from the Thunderbolt signals is already handled by existing adapters in the market.

    By having both on a single port it simplifies docking, especially when used with a monitor that has a USB 3.0 hub integrated.

    The extra chip is a temporary situation. It wasn't that long ago that having USB 3.0 in a system meant an extra chip on the motherboard. Now it's part of core chip set, which in turn is increasing integrated to the CPU.

    Millions of machines have eSATA ports but only a fraction of them are ever used. Just pointing it out can be a good way to get a blank star from a lot of people. But it doesn't add much cost and has great value for those who need it. How many PCs even need more than two SATA ports for the hard drive and optical drive? Many business desktops are so limited. But would you buy a motherboard today for a full size machine that didn't have at least six ports available? Even though you might never use more than half of them?

    When you're measuring in nickels and dimes across millions of units, you have to make educated guesses over what is simply a waste and what is a valued feature, even if it goes largely unused by a large portion of the consumers.

    Once the premium falls within a certain cost envelope you'll see nearly every new machine equipped for Thunderbolt. By our presence in this forum, we're the sort who tend to be a bit ahead of the curve and likely to discuss a technology before it is ready to be mainstream.
    Reply
  • AggressorPrime - Tuesday, April 09, 2013 - link

    No progress since Thunderbolt first launched. Just an optimization that should have been there since the beginning. That said, if I am running a DP 1.2 signal on the up-link channel, can I split the 20Gbps down-link channel into a 10Gbps bidirectional channel and use that to manage a device that wants back and forth communication?

    Also, can I have 2 up-link channels running at 20Gbps, like to run 2 4K monitors from a single cable?
    Reply
  • repoman27 - Wednesday, April 10, 2013 - link

    The only thing that has apparently not progressed one iota since the launch of Thunderbolt is the understanding of the technology by most of those posting comments about it online.

    Why would Intel have included DisplayPort 1.2 support in Thunderbolt controllers before they included it in the processor platforms they were designed to accompany? How could Intel have improved the power characteristics of their first gen controllers without performing these generational iterations and moving to progressively more efficient fabrication processes? You may as well say that Haswell should have been there since Sandy Bridge.

    That said, I believe the way it will work is that Falcon Ridge devices can connect to each other as one 20 Gbit/s, full-duplex channel, or connect to a previous generation controller via 2 separate 10 Gbit/s, full-duplex channels. Each direction in a channel can carry DisplayPort and/or PCIe packets. Each controller has a crossbar switch, so you wouldn't actually be splitting the channel, you'd just be using the switch to share the channel's bandwidth between devices.

    It is unlikely that Falcon Ridge will offer 2x 20 Gbit/s channels, and even more unlikely that Intel would ever offer a simplex or half-duplex Thunderbolt arrangement. Since there aren't any 4K Thunderbolt displays yet, running 2 would not be possible until such time as they become available (and one could afford them) anyway. You could run two 4K displays via a single DP 1.2 link, just not at 24 BPP, 60 Hz. Or you could simply use both ports of a host with two ports.
    Reply
  • ShieTar - Wednesday, April 10, 2013 - link

    So, how many more generations until can start putting the GPU into the monitor? Reply
  • kwrzesien - Wednesday, April 10, 2013 - link

    Now we're talking! Reply
  • DanNeely - Wednesday, April 10, 2013 - link

    Ugh. If you did that you'd need to throw out your monitor every time you upgraded your GPU. Even a cheap crappy monitor costs at least $100; a good one will be at least that. A a stroke you've doubled the cost of a GPU upgrade. Reply
  • ShieTar - Thursday, April 11, 2013 - link

    No you would not. You would have to remove the old GPU from the monitor and place the new in. I see no technical reason not to make this easily achievable.

    The point is just, that at some point with increasing resolutions and refresh frequencies, we are likely to come to a point where we need to transfer much more data from the GPU to the screen than from the CPU to the GPU. And at the same time, we will build CPUs which are good enough for GPU-restricted tasks into more and more of our devices. So by moving the GPU into the screen, you can connect your Laptop or your mobile or your TV-Receiver to your screen, and it is always capable of 8K resolution at 200Hz (One might dream).
    Reply
  • epobirs - Tuesday, April 16, 2013 - link

    High-end video cards are not small items. I wouldn't want to bulk up my monitor with such a thing. A small box inline between the user device and the display should serve just fine. It would be another role for a good docking station that takes your highly portable device and instantly turns it into a powerhouse desktop system.

    On the one hand, that means no need to sync data between a desktop and a portable. On the other, it means losing the use of both modes if you lose the portable portion.
    Reply
  • Alien959 - Wednesday, April 10, 2013 - link

    What about upgrading notebook graphics? I have an older Clevo laptop with core 2 duo T7250 but with integrated S3 Graphics GPU if something like thunderbolt was around then probably it can bring new life in that notebook with external gpu. Reply
  • nickeditor - Thursday, April 11, 2013 - link

    Is not clear if current computers with Thunderbolt can bump up to 20Gb/s
    New chip implies new computers?
    Are they talking about Thunderbolt 2.0?
    Reply

Log in

Don't have an account? Sign up now