Bandwidths and approval years are:Thunderbolt didn't cripple anything. Thunderbolt doesn't support the DisplayPort 1.2 HBR2 link rate, but it does support multiple streams of DisplayPort 1.1a and supplies far more aggregate video bandwidth than any version of HDMI to date. Considering that the first sink device to support DisplayPort 1.2 was only certified on August 22, 2011, and I haven't seen any consumer products that yet require it, I don't believe many folks will notice the lack of DP 1.2 support for some time yet.
hdmi 1.0, 2002, 3.96 Gb/s
hdmi 1.3, 2006, 8.16 Gb/s
dp 1.0/1.1, 2006/2007, 8.64 Gb/s
dp 1.2, 2009, 17.28 Gb/s
tb's dp, 2011, 8.64 Gb/s
So hdmi & tb are quite the same, but tb was a step back from dp.
As for future use, it might get sooner than many of us think.
Since ipad now has over 3Mpx display and Apple will be shipping HiDPI version of OsX, we could very likely get over 10Mpx displays within a year.
Eg. 3840x2400x60Hz@10bit would need 16.59 Gb/s.
Well, if Cactus Ridge raises tb's dp back to v1.2...
Of course technically tb is superior, but if it will be too expensive to breakthrough to mainstream, there will be much more limited variety of products using it and at the same time, same products with usb3 and a bit lower specs, but with a fraction in price.The biggest problem with ExpressCard slots is the insane amount of real estate they require. They are enormous by MacBook Air / Ultrabook standards. Apple dropped ExpressCard slots on all but the 17-inch MacBook Pros because there was no way to justify the amount of space they consume in smaller devices, especially once they switched to non-removeable batteries. Thunderbolt provides twice the PCIe bandwidth of, and more data bandwidth altogether than ExpressCard 2.0, which has barely taken off yet. Furthermore, it sits on top of a small and already present port. By also allowing daisy-chaining, Thunderbolt is much more flexible than giving users just a single ExpressCard slot.
About EC's real estate, it isn't so insanely bigger if you compare it to combination of sd card slot, tb connector and tb controller.
But if you think that Apple gave us tb instead of usb3 for better computing experience and not just to rip us off with fancy name & $50 cables, can you explain why did they took EC readers off?
Because mainstream users didn't use it?
Well, they won't use tb either...
Did unibody as construction eat all space up, so it didn't fit there anymore?
Did internal battery needed so much more space, that EC didn't fit anymore?
And the biggest question: why did Apple thought 2.5 years that MBP does not need EC, but now thinks that all computers need TB?
Imacs or minis didn't need any expandability, although there has always been space for EC slot, nor did they need any other fast connection (esata, usb3, 10GbaseT), but now they need it?
You couldn't daisy chain EC cards, but there are still problems with tb.
Major problem comes from that you have only 1 port that you share between display data and other data.
Within first year of existence there is only 1 single model of displays, which supports tb. Monitor manufacturers do not seem to trust that tb will succeed...
You can't attach 2 dp monitors to tb port (even if controller could feed 4 displays?!?). Or you can't have 1 dp monitor and 1 tb device with only 1 tb port. (Hmmm, at least half of different tb devices on market now has only 1 port...?)
So are you surprised with this 5.7%?The 5.7% figure is based on the total number of USB 3.0 controllers shipped by all foundries to date, and the estimated total number of PC's currently in use in the world today. It is on the high side because there are clearly many motherboards, add-in cards and PC's out there that contain multiple USB 3.0 controllers. Also there are quite a few that were used on Intel boards that were scrapped because of the 6-series SATA bug, and many that are still in the supply chain that aren't being used by anyone yet.
My crystal ball is a bit foggy at this point when it comes to Thunderbolt. I guess I don't see more than a 20-25% attach rate by 2017, and maybe about 10-20% user adoption rate. So yeah, less than 5% for ThunderBolt. I think 100% user adoption rate of USB 3.0 is optimistic though. Many classes of devices will have no need to progress beyond USB 2.0 in the next 5 years. Those users that don't require the types of devices that will inevitably shift to USB 3.0 may be plugging something into a USB 3.0 port, but it won't be a USB 3.0 device. You could argue that anyone who plugs a DisplayPort device into a Thunderbolt port is a Thunderbolt user, in which case usage figures might be much higher.
Many people will no doubt be happy with USB 3.0's 400 MB/s, but there is no avoiding the fact that a Thunderbolt cable can carry more than 5x that much today and with much lower latency. Thunderbolt is what you need when you need more than 400 MB/s. Even if most people don't, for those that do, the choice will be obvious.
Most computers never get upgraded. So only way to get saturation up, is to wait that old computers die away.
I'd bet that 99% of computers made next year, that have usb ports, will have usb3. Then you wait that installed base renews, about 5 years and usb3 will have 99% market penetration and there's just nothing that could prevent this happening. (Okay, if Apple keeps their stubborn attitude, only 95%...)
As for tb, it might be just another ADC or fw. It can die away fast or just remain alive in niche, in co-existence with eg. 10GbaseT, usb4, HDbaseT, hdmi1.5, etc...
Last edited: