Thunderbolt Performance

The Eagle Ridge Thunderbolt controller is home to two Thunderbolt channels, each one is good for up to 10Gbps in either direction (up or downstream). That works out to be 20Gbps of bandwidth per channel or 40Gbps aggregate between the two. You can only send two channels worth of data down a single Thunderbolt cable, so there's no point to having more than two from a performance standpoint unless you have more than one port on your system.

If DisplayPort and PCIe traffic are indeed carried on separate channels, then the Thunderbolt Display by itself is eating up around 70% of the bandwidth of a single channel on its own (2560 x 1440 x 32bpp x 60Hz with 8b/10b encoding > 6.75Gbps). That leaves 10Gbps in each direction for PCIe traffic. If we look at our benchmarks from the previous section we see that we can hit just under 2Gbps with all of the auxiliary interfaces (GigE, FW800, USB2) running. Given our previous investigation with the Promise Pegasus we know that 8Gbps is feasible there as well. It's possible, that with everything running at once, we could actually run into bottlenecks with Thunderbolt.

To find out I ran a few tests. First I needed a baseline so I threw four SF-2281 SSDs into the Pegasus R6 chassis and configured them in a RAID-0 array. I ran a 2MB sequential read test (QD=16) and measured 909MB/s from the array. This value was obtained without the Thunderbolt Display connected, only the Pegasus R6.

Next I connected the Thunderbolt Display directly to my test MacBook Pro, and then connected the Pegasus to it. I repeated the test, this time getting 900MB/s. Thankfully the presence of the Thunderbolt Display doesn't seem to impact the max data rate I can get from the Pegasus.

For my third test I added a Gigabit Ethernet transfer from a file server to a local SSD using the GigE port on the display. During this test I was also playing back music using the Thunderbolt Display's internal audio codec and speakers. I re-ran the Pegasus test and got 855MB/s.

For my final test I re-ran the third test but added a FireWire 800 to USB 2.0 SSD transfer, both connected to the Thunderbolt Display. I also fired up the FaceTime HD camera on the display using Photo Booth and left it on during the test. The final performance score from the Pegasus was 817MB/s.

Apple Thunderbolt Display Performance

With everything running Thunderbolt performance took a 10% hit. Note that the standard Pegasus configuration isn't able to hit these data rates to begin with, so unless you've pulled out the 12TB of storage and stuck in your own SSDs you won't see any performance drop.

What this does tell me however is the ultra high end users that are looking to daisy chain multiple Thunderbolt storage boxes together may not want to do so. I only have a single Pegasus R6 on hand, but I'm guessing there will be significant performance drop off after the first box. Not that I'm complaining about being able to push nearly 1GB/s over a $49 cable from a notebook, I'm just trying to give a heads up to those who may have aspirations of even higher performance.

Testing the Pieces Display Testing - Color Quality & Uniformity


View All Comments

  • repoman27 - Tuesday, September 27, 2011 - link

    I'm just going off the information available, like everyone else. Which is why probably half the statements I've made about TB thus far have turned out to be inaccurate.

    But the TB port is electrically equivalent to a mini DP port. There are no additional contacts. A TB port operates in either DP compatibility mode or in TB signaling mode. The 4 pairs of signaling wires are used for either 1 DP main link Tx or 2 TB bi-directional channels, both of which are capable of transporting (tunneling) a DP main link. As far as I can tell at least.

    I posted a block diagram earlier, which I'll post again here, although I need to clean it up a little as my current theory of things Thunderbolt is now a bit more plausible.
  • repoman27 - Monday, September 26, 2011 - link

    I just noticed that my math was a little too quick and dirty. There's only one FireWire 800 port on the ATD, not two. So actually a second USB 2.0 controller or memory card reader were not out of the question at all, and probably just failed to make the cut due to cost.

    An optical drive would have required a SATA controller, which wouldn't have fit within the bandwidth limitations, and altogether would have added about another $50 to the price tag.
  • repoman27 - Sunday, September 25, 2011 - link

    I find it bizarre the number of tinfoil hat conspiracy theorists that post comments on every review of an Apple product saying that the review was biased or nothing but a paid advertisement. I’m not sure how you could read this article and not believe that it was anything other than the most objective, informative and detailed review of the ATD yet published.

    Some points for consideration:

    There are plenty of reviews on Anandtech for products made by manufacturers that also advertise on Anandtech, yet these are never called into question. It’s just Apple, who to my knowledge have never actually advertised on a review site.

    While Apple does submit products for review, Anand went out and additionally bought his own ATD to dissect, so he actually paid Apple this time around. I doubt his review unit came packed in a box full of cash.

    There are no generally negative reviews posted on sites like these because the authors tend to spend their time reviewing products that have merit. Most of the major manufacturers don’t make a habit of bringing to market products that are clearly defective or devoid of any useful functionality. Apple makes solid products, and therefore gets a lot of positive reviews. Not that they are by any means above critique, but their standards tend to be higher than many of their competitors.

    Being objective is not the same as being impartial. The biases of reviewers will always show up in what they write, but stating opinions does not necessarily diminish the objectivity of the data presented.

    People who can’t figure out the timing of the posts on this site are baffling to me. When AMD, Intel or NVIDIA release new products, we get a quick blurb when it is announced and a full review a few days after it ships. When MWC is going on you get a ton of mobile device news, when IDF is going on you see a mess of Intel related posts, when BUILD is happening you’ll have a bunch of Microsoft stuff to check out. Yet so many people submit comments saying that this site has jumped the rails when WWDC results in a week of Apple posts.

    Why did the ATD get an in depth review? Because it marks the release of a highly anticipated piece of tech which was surrounded by many unknowns due to the fact that it actually forges into some new territory. Is there really a good reason why this site would NOT review this device? The only reason I can think of would be if a competitor had paid them off.

    If all those bitching about the quality or integrity of this site can point to better, more objective, more in depth reviews of these products, please feel free to cite references.
  • Brigga - Sunday, September 25, 2011 - link

    As simple and probably silly as this sounds why doesn't apple have a docking station? Port replicators suck, but docking stations can be quite unobtrusive and easy to use (ie. no disconnecting of cables and second power supply stays where it is). Reply
  • repoman27 - Monday, September 26, 2011 - link

    Yeah, actually Apple solved the problem that way for five years back in the early nineties. The PowerBook Duo line had a docking port that worked with the Duo Dock and several other docking accessories. I'm guessing that they have enough bad memories of the problems inherent with those designs that they're shooting for a slightly different approach now. Reply
  • pesos - Sunday, September 25, 2011 - link

    I went out and picked up a TB display last night to pair with my 2011 mbp. I live in windows about 95% of the time and expected there to be no issues with the basic usb hub built into the display – unfortunately that does not seem to be the case. My Apple usb keyboard (also brand new) and dell usb mouse both hang/freeze/work erratically when connected to the back of the display, as does my Jabra pro headset. I can make the behavior 10 times worse simply by using hardware built into the display, such as the facetime camera. If I fire up Lync and start a video chat, my keyboard and mouse become nearly unusable. My basic testing under Lion appears to confirm this is an issue with Apple’s boot camp drivers as things work fine under OS X. Problem occurs both with win7 and win2008r2. Reply
  • shin0bi272 - Sunday, September 25, 2011 - link

    first world problems... Reply
  • Oli56 - Sunday, September 25, 2011 - link

    Anand, great review but it seems you completely overlooked the impact of the glass which can be negative to some users. I owned an iMac some time ago and the glass reflections were very disturbing to me and led to ocular fatigue. It is unfortunate Apple has moved to glass display for all its line of screens, fortunately we can still have a MacBook Pro with mate display as option but that doesn't exist for a 27" IPS screen which should really benefit from it. So basically glass maybe looks cool on display but when this is about to work all day with it, this is a nightmare. This is the only thing keeping me away from that screen, as the rest of it looks great. Reply
  • dmaffei - Monday, September 26, 2011 - link

    You never fail to impress... nice job old friend!!!

    Dan :-)
  • StefanoT - Monday, September 26, 2011 - link

    so this really, really, is an Apple Display without DisplayPort?!
    If I'm not mistaken, just the other day DisplayPort was the way to the future, touted to eventually supersede DVI/HDMI. Indeed, it was almost too new to be widely supported! DisplayPort-enabled monitors, in my part of the world, are still few and far between. I was still waiting for the DisplayPort goodness to free me from carrying all sorts of miniDP adapters. To DVI, to HDMI, even (gosh!) to VGA!
    And now, suddenly, DisplayPort is to fall into oblivion, no longer supported by Apple, which was one of the main proponents and, ironically, now supports a standard that is built-on and, in a way, includes DisplayPort, and on top of that shares the same connector.
    Anand points out that the company has never got emotional about legacy stuff. If they did, the new technology would never get adopted. Apple is leaping to the future and can't get stop to deal with yesterday's interfaces. I would say that between one leap and another, you need to "land" somewhere. I'm not a fan of never-ending legacy support, but I'm still waiting for VGA to disappear from mainstream netbooks. Are they suggesting that I either ditch my oh-so-non-2011 devices (i.e., "not-released-in-2011", as opposed to "not-bought-in-2011") in order to use their new display, or that I buy an old, "non-Thunderbolt" version of their display, which sells for the same price? This doesn't make sense any way you put it... and to think that they had it so easy with Thunderbolt. The thing seems built with backward compatibility in mind.
    I didn't get if future displays will eventually support both DisplayPort and Thunderbolt. Perhaps once Intel makes it piece-of-cake easy, as opposed to "just add a chip to that big motherboard you've got back there, behind the glass". Perhaps once other companies start offering it. Or perhaps never, because why should Apple not leapfrog everybody again (including its customers), by switching to an all-new technology next year?

Log in

Don't have an account? Sign up now