Thunderbolt 3: Mismatch and Opportunity
After the relative failures of the Thunderbolt 1 and Thunderbolt 2 cabling systems, Thunderbolt 3 appears set for success as the “one cable to rule them all.” Thunderbolt 3 is billed as a user-friendly cabling system, robust enough to support everything from high-definition video to high-res audio to personal storage networking. It operates at 40 Gbps, twice the data rate of earlier Thunderbolts. Thunderbolt 3 also supports the most popular consumer electronics interfaces including the USB Type-C connector, USB 3.1, DisplayPort 1.2 and HDMI 2.0. As an aside, some insiders are now talking about an 80 Gbps Thunderbolt 4 at some time in the not-too-distant future.
The proof that Thunderbolt 3 is catching on is that many important laptop OEMs are now adopting Thunderbolt 3. In the past the Thunderbolt reality has been an Intel-Apple one. Thunderbolt was born at Intel at the end of the last decade, but despite Intel’s early high hopes, the only big computer/consumer electronics firm to have implemented the earlier versions of Thunderbolt was Apple. Thunderbolt 1 and Thunderbolt 2 did not see much (or indeed any) love at the Windows/PC companies. But this has changed with the advent of Thunderbolt 3. Acer, Asus, Dell, HP, and Lenovo began to show real interested in Thunderbolt in 2015 and all have high-profile Thunderbolt 3 capable laptop models on the market today.
Meanwhile, Apple has gradually strategically positioned Thunderbolt to be its key interface. In some Apple products, the Thunderbolt port is the only port in the box.
CIR’s analysis, however, suggests that there is a mismatch between the Thunderbolt’s claim to be a multi-application cabling platform and the technology that has been central to all the previous Thunderbolts: active copper cabling. The mismatch is that the standard active copper cabling for Thunderbolt 3 cannot fully support Thunderbolt’s broad ambitions with regard to applications.
As a result, we think that Thunderbolt 3 is now creating an opportunity for mass market consumer optical networking. In any case, the fact that at a top speed of 40 Gbps, Thunderbolt 3 operates at the same data rate that medium-sized data centers are using for interlinks, seems a prima facie case that an optical implementation makes sense.
Recent estimates in CIR’s report, “Markets for Active Optical Cables: 2017-2026,” suggest that optical Thunderbolt cables will generate $620 million in revenues by 2022. However, that number is highly dependent of how quickly Thunderbolt 3 cable is commoditized.
The Need for Optical Thunderbolt
To illustrate what is currently expected of Thunderbolt 3, consider that the Thunderbolt 3 hardware reportedly provides enough speed to deliver video to two 4K displays at a 60 frame per second refresh rate or to a single 4K display at 120 frames per second. But not only is such a set up nothing out of the ordinary these days, but bandwidth demands on video networking platforms are growing, driving the need for even more speed:
- Apple has offered 5K monitors for a couple of years now and at the just-completed Consumer Electronics Show (CES) in Las Vegas, I saw the very impressive 8K machines: these will reach the market in the not too distant future. Such displays would not just be platforms for advanced television, but would be used for editing of home videos captured with a cell phone
- Video will not be the only bandwidth hungry application that will find Thunderbolt useful in the future. Gaming applications – especially those based around virtual reality (VR) – will eat up bandwidth too -- VR was also very much a “thing” at CES. More speculatively, telepresence and autostereoscopic 3D might find a need for Thunderbolt 3 or even a future Thunderbolt 4.
Thunderbolt 3’s distance problem: With active copper cabling Thunderbolt 3 and its earlier versions, standard data rates can only be maintained error free up to three meters. Using standard copper implementations of Thunderbolt – networking over three meters can lead to unacceptable bit error rates (BERs) and/or low-quality video. Future implementations of Thunderbolt on copper may be even more restrictive in terms of the distance they could carry a signal over copper.
If the extent of a network is a desktop, three meters is fine. But consider a home office or home theater network. If a cable needs to wind its way around a sofa or some other obstacle, or even into the next room, three meters of active copper cabling won’t do. Future applications such as shared virtual reality are certain to need more than 3 meters. Home theater, personal storage applications and perhaps personal video editing need it now. In the Exhibit below, we spell out more closely where a strong use case can be made for optical Thunderbolt 3.
Current optical versions of Thunderbolt cable (Corning is the mindshare leader here) extend the distance limit to 60 meters, which is acceptable for most applications in which Thunderbolt might be used. There are also other drivers for optical Thunderbolt. While fiber optics is often characterized as more expensive than copper, the reality today is that this is more matter of relatively small numbers of fiber optics links that are produced, not inherent cost factors. High volume production of optical Thunderbolt 3 cabling could put it in a very favorable competitive position when compared with copper Thunderbolt 3.
How Thunderbolt Promotes “Mass Market” Fiber Optics
Thunderbolt 3 could therefore prove to be the first mass market optical interface and at prices that easily match copper. CIR believes that there are two reasons for believing this will in fact happen:
- Thunderbolt 3 (copper and optical) will become ubiquitous, because of its multivendor and multi-application support. This in turn will create a large addressable market for optical versions of Thunderbolt 3.
- Second, inherent distance limitations for copper Thunderbolt cables, implies that optical implementations will penetrate that addressable market both quickly and deeply. This is because – as we have seen – there are many applications (actual and potential) where copper Thunderbolt just won’t cut it.
The bottom line therefore is that Thunderbolt 3 portends to be a new direction for consumer datacom that will be everywhere and will need fiber optics to make it happen.