Hey, Whatever Happened To superMHL?

There is no such thing as a ‘sure thing.’ You can have a 20-yard field goal try with 5 seconds left, two foul shots left to ice the game, or a one-on-one penalty shot with your best wing on the ice. Doesn’t matter – things do go awry. In fact, sometimes they never get going in the first place.

Two years ago this coming January, Silicon Image (now Lattice Semiconductor) unveiled what they claimed to be the best next-generation display interface. They called it superMHL, and it was super indeed; sporting a large, 32-pin symmetrical plug design to go with a 36 gigabits-per-second (Gb/s) data transfer rate.

That’s wasn’t all. superMHL (basically MHL on steroids) also supported the new Display Stream Compression (DSC) 1.1 standard. And it would also work with the all-important USB 3.0 Type-C plug’s Alternate Mode, which multiplexed display connections and fast USB serial data in the same ‘smart’ plug.

Wow! I didn’t see this coming; neither did most of the trade press in attendance. Here was a connector faster than DisplayPort’s version 1.3 (32 Gb/s), plus it was symmetrical in operation (plug it in either way, it doesn’t care, it’s smart enough to set itself up the right way). And it was compatible with the next generation of USB connectors.

Even more amazing, the MHL Consortium demo showed 8K content flowing to a large Samsung 8K TV through this interface, which claimed to support 7680×4320 video @ 60 Hz with 4:2:0 color (albeit using DSC to pack things down a bit in size). If there was ever a ‘sure thing,’ this was it!

It's the fastest display interface out there - and no one uses it. Maybe they should call it HDMI 3.0?

It’s the fastest display interface out there – and no one uses it. Maybe they should call it HDMI 3.0?

I was assured in the following months that Lattice and the MHL Consortium would have several press announcements pertaining to design wins for the 2015 holiday season. I’d see several new UHDTV televisions with at least one superMHL port and the rest of the inputs would be HDMI 2.0 connections. Thus, we’d be ready for the brave new world of 8K TV! (Never mind that 4K TV was still getting on its feet at the time!)

But it never happened. Black Friday, Christmas, New Year’s, and then ICES and the 2016 Super Bowl came and went with no announcements. At ICES 2016, the MHL Consortium once again had a demo of 8K content playback through an LG 98-inch LCD TV using the superMHL interface, and “yes, it looked great” and “we’re ready for 8K TV” and “it works with USB Type-C” and so on, and so forth.

Right now, it’s pretty much radio silence about superMHL. So what happened?

For one thing, the adoption rate of HDMI 2.0 since its formal unveiling in 2013 can be charitably described as “slow.” Early Ultra HDTVs had perhaps one HDMI 2.0 port on them, and not all of them supported the new HDCP 2.2 copy protection protocol. In our industry, we’re only now starting to see distribution amplifiers and switches with HDMI 2.0 connections – there’s still a lot of version 1.4 product out there, too.

Another perplexing question: Since superMHL fixes the speed limit problems of HDMI 2.0 by doubling them – and also adds the all-important compatibility with USB Type-C (a must, going forward) along with support for DSC (critical as we push display resolutions beyond 5K), why would Lattice continue to support both formats, or even suggest they could be mixed on future UHD+ televisions and monitors?

In other words; if there is a better option, then why wouldn’t you want that option?

To be sure; Lattice is in a tricky position. Through their subsidiary HDMI Licensing LLC, they reap millions of dollars each year in royalties associated with every HDMI port on every piece of consumer and commercial gear. That’s a nice cash flow, and who wants to mess with it?

But they really can’t lose here, inasmuch as they control the IP for all of these transition-minimized differential signaling (TMDS) interfaces. Why not bite the bullet and announce the phase-out of HDMI 1.3/1.4, and move everyone to version 2.0? Better yet; just announce a sunset for version 2.0 and start the transition to superMHL, a/k/a HDMI 3.0?

Yeah, it's fun to demo 8K TV using superMHL, but that takes the focus off the real-world, practical interfacing solutions we're facing now.

Yeah, it’s fun to demo 8K TV using superMHL, but that takes the focus off the real-world, practical interfacing solutions we’re facing now.

One problem Lattice created with this new connector is that it’s effectively an oxymoron. MHL stands for Mobile High-definition Link, and it was originally designed to multiplex HDMI signals through 5-pin micro USB ports. The concept was that the single micro USB connector on your smartphone or tablet could connect to a television so you could play back videos, show photos, and share your screen. (Never mind that the majority of people prefer to do this via a wireless connection and not a 15-foot HDMI-to-micro USB cable that often requires a power adapter.)

So MHL meant “small, fast, and powerful.” And now we have the ‘funny car’ of display interfaces with a large connector that will never get anywhere near your mobile device…and the way things are going, it may never get anywhere near your TV, either.

In previous columns and in my classes and talks, I’ve written about the deficiencies of HDMI 2.0 – slow speed, non-symmetrical, no support for USB Type-C (finally remedied a few months ago) and lack of support for Display Stream Compression. superMHL fixes all of these problems in one fell swoop.

The answer? Re-brand this connector as HDMI 3.0 – which it really is – and make the appropriate announcement in two months at ICES 2017. Practically speaking; MHL has been a non-starter (among major U.S. brands, only Sony, Samsung, and LG have supported it on their smartphones and TVs) and the adoption rate for HDMI 2.0 is nowhere near as fast as it was for version 1.3. Too many interfaces and too much confusion!

After all, even Elvis Presley had to make a comeback…