NVIDIA’S Tegra 3 Leading the Way for Smartphone Modularity

I have been an advocate of modularity before it became popular to do so. The theory seems straight-forward to me, in that if the capabilities of a smartphone were outpacing the usage model drivers of a rich client PC, then consumers someday could use their own smartphone as a PC.  Large displays, keyboards and mice still exist in this usage model, but the primary intelligence is in the smartphone then combined with wireless peripherals.  At this year’s Mobile World Congress, NVIDIA took us one step closer to this reality with their partners and the formal announcement of Tegra 3 based smartphones.

Tegra 3 for Smartphones

Tegra 3 is NVIDIA’s latest and greatest SOC for smartphones, “superphones“,  and tablets.  It has four ARM A9- based high performance, 1.5 GHz cores and one “battery saver” core that operates when the lowest power is required.  The fifth core comes in handy when the system is idling or when the phone is checking for messages.  Tegra 3 also includes a very high performance graphics subsystem for games and watching HD video, much more powerful than Qualcomm’s current Adreno 2XX hardware and software implementation.

clip_image004NVIDIA announced five major Tegra 3 designs at Mobile World Congress; the HTC One X, LG Optimus 4X HD, ZTE Era, Fujitsu’s “ultra high spec smartphone” and the K-Touch Treasure V8.  These wins were in what NVIDIA coins as “superphones” as they have the largest screens, the highest resolutions, the best audio, etc.  You get the idea.  For example, the HTC One X sports a 4.7″ 720P HD display, the latest Android 4.0 OS, Beats audio, NFC (Near Field Communication), and its own image processor with a 28mm lens to take great pictures at extremely low light.  You get the idea.

There is a lot of goodness in the package, but that doesn’t remove the challenge of communicating the benefits of four cores on a 5 inch screen device.

Quad Core Phone Challenge

As I wrote previously, NVIDIA needs to overcome the challenge of leveraging four cores beyond the spec on the retail tear clip_image002pad.  It’s a two part challenge, the first to actually make sure there is a real benefit, then to articulately and simply communicate that.  These are similar challenges PC manufacturers had to deal with.  The difference is that PC makers had 20 years of dual socket machines to establish an ecosystem and a messaging system.  Quad core tablets are an easier challenge and quad core convertibles are even easier in that you can readily spot places where 4 cores matter like web browsing and multitasking. Smartphones is a different situation in that due to screen size limitations, multitab browsing and multitasking rarely pegs a phone to its limits.  One major exception is in a modular environment where NVIDIA shines the most.

Tegra 3 Shines the Most in Modular Usage Models

Modularity, simply put, is extending the smartphone beyond the built-in limitations. Those limitations are in the display, audio, and input mechanisms.  When the smartphone breaks the barriers of itself, this is where NVIDIA Tegra 3 shines the most.  I want to be clear; Tegra 3 is a competitive and differentiated smartphone and tablet SOC without modularity, but is most differentiated when it breaks free from its limited environment.

NVIDIA has done a much better job showing the vision of modularity but its partners could do a better job actually delivering it.  On the positive side, partners are showing some levels of modularity. HTC just announced the HTC Link for the HTC One X, software and hardware solution that plugs into an HDTV where you can wirelessly mirror what is on the phone’s display.  It’s like Apple’s AirPlay but better in some ways like being able to project a video on the large display and do something different on the phone display, like surfing the web.  Details are a bit sketchy specifically for the HTC One X and HTC Link, but I am hopeful they will roll out some useful modular features in the future for usage models. Apple already supports wireless mirroring supporting games so in this way, HTC Link is behind.

What NVIDIA Tegra 3 Should Do

What NVIDIA’s partners need to create is a game console and digital media adapter solution that eliminates the need to buy an XBOX, PlayStation, Wii, Roku, or Apple TV.  The partners then need to attack that.  All of the base clip_image006software and hardware is already there and what HTC, ZTE, or LG needs to do now is package it to make it more convenient for gaming. This Tegra 3 “phone-console” should have a simple base near the TV providing it power, wired LAN, HDMI, and USB.  This way, someone could connect a wireless game controller and play games like the recently announced Tegra 3 optimized games in great resolutions with rich audio. The user would have the ability to send phone calls to voice mail or even to a Bluetooth headset.  Notifications can be muted if desired as well.  And of course, if you want to watch Netflix, Hulu, or Amazon movies it’s all there, too.  The alternative to this scenario is for a Wi-Fi Direct implementation that doesn’t require a base where the user can utilize the phone as a multi-axis game controller with force feedback.  The challenge here is battery life but the user can pause the game or movie and pick up phone calls and messages. This usage model isn’t for everyone, but think for a moment about a teenager or college bound guy who loves gaming, wants a cool phone, and doesn’t have the cash to buy everything.  You know the type.

Other types of modularity that NVIDIA’s partners must develop are around productivity, where the phone drives a laptop shell, similar to Motorola’s Lapdock implementations as I analyzed here. Neither the software, hardware, or price made the Lapdock a good solution, but many of the technologies now exist to change that.  NVIDIA’s Tegra 3 would be a great start in that it enables real multitasking when using the Lapdock in clamshell PC mode.  Android 4.0 provides a much more modular computing environment to properly display applications on a 5″ and 11″ display including scaling the fonts and reorienting windows.  The Motorola Lapdock used two environments, one Android Gingerbread a a different one for PC mode.  Unsurprisingly, it was a good start but very rough one too, with room to improve.

NVIDIA, the Silicon Modularity Leader with Tegra 3

NVIDIA with its Tegra 3 solution is clearly the current silicon leader to support future modular use cases.  They are ahead of the pack with their modularity vision, patiently waiting for their partners to catch up.  This was the most evident at CES where NVIDIA showed an ASUS Transformer Prime connected to an XBOX controller and an HDTV playing high quality games. They also demoed the Prime playing high end PC games through remote desktop. Now that is different.

The opportunity for HTC, ZTE, LG and potentially new customers like Sony, RIM, and Nokia is there, and the only question remains is if they see the future well enough to capitalize on it.  With all the complaints from handset vendors on differentiation and profitability with Android, I continue to be puzzled by their lack of aggression.  An aggressive handset maker will jump on this opportunity in the next two years and make a lot of money doing in the process.

Why the PC Industry Cannot Ignore Smartphones

When HP abandoned their smartphone and tablet business and webOS last August, many in the industry were hp-veerdisappointed in the speed of the Palm acquisition and the quick dismantling of it. Some who consider themselves "business-savvy" said it was the wise approach as it wasn’t core to HP’s corporate mission. They said that smartphones were a distraction to competing with IBM and even Dell. We won’t know until 3-5 years from now whether it was a good decision or not.

I believe though, that just as PC companies fought to stay away from the sub-$1,000 PC market in the 90’s, PC makers who don’t embrace smartphones could be out of the client hardware business in 5 years.

Some Context

Over the last 20 years, PC hardware and software have done this little dance where one is ahead of the other. New software came out that required better hardware, then the new hardware outpaced the old software and the cycle continued. With the better hardware and software came new features and usage models like multimedia, desktop publishing, 3D games, DVD video, videoconferencing, digital photography, the visual internet, and video editing. Then Microsoft Vista was launched and it seemed no matter how much hardware users threw at it, issues still existed. Microsoft then spent the next few years fixing Vista and launched Windows 7 instead of developing environments for new rich client usage models. Windows 7 actually took less hardware resources than Vista, the first time a Microsoft OS could say this. Microsoft is even publicly communicating that Windows 8 will take less resources than Windows 7. So what happened? Did the industry run out of usage models to consume rich PC cycles? No, there are many usage models that need to be developed that use rich PC clients.

What happened was netbooks, smartphones and tablets. Netbooks threatened Microsoft and forced them to re-configure Windows XP for the the small, cheap laptops. This was in response to the first netbooks, loaded with Linux, getting shipped into Best Buy and direct on the internet. In retrospect this wasn’t a threat to Microsoft, as those netbooks had a reported 50%+ return rate. After netbooks came MIDs and after MIDs failed came touch smartphones and the iPad. Once the iPhone and iPad showed strong sales it was clear that the center of design was moving to mobility even though needs the rich client PC could solve didn’t just go away.

Windows 8 and Rich PC Clients

Windows 8 was clearly architected to provide a tablet alternative to the iPad and stem the flow from Windows to iOS and Android. Most of the work has been to provide a new user and development environment called Metro, WinRT and to enable ARM SOCs. None of these investments does a single thing to propel the traditional rich PC client forward, maybe with the exception of enabling touch on an all-in-one desktop. Without Microsoft making major investments to propel the rich client forward, it won’t move forward even to the dismay of Intel, AMD and Nvidia. I want to be clear that there are still problems that the rich client PC can solve but the software ecosystem and VC investment is enamored primarily with tablet, smartphones and the cloud. Without Microsoft’s investment in rich PC clients, thinner clients like phones and tablets will evolve at a much faster rate than rich PCs.

The Consequences of Not Investing in the Rich PC Client

With the software ecosystem driving "thin" clients at a much faster rate than "rich" clients, the consequences start to airplaytvemerge. We are seeing them around us every today. Users are spending more time with their tablets and smartphones than they are with their PCs. Savvy users are doing higher-order content creation like photo editing, video editing and even making music with GarageBand. That doesn’t mean that they don’t need their PCs today. They do, because neither smartphones nor tablets can do everything what a PC or Mac can do…. at least today. Display size, input method and lack of software modulraity are the biggest challenges today.

Enter Smartphone Modularity

Today, many users in traditional regions require at least a smartphone and a PC, and a tablet is an adder. Tomorrow, if users can easily attach a keyboard to a tablet via a convertible design, they may not need a PC as we know it today. It’s not a productive discussion if we debate if we call this a PC with a removable display or a tablet with a keyboard. What’s important is that some users won’t need three devices, they’ll just need two.

What about having just one compute device, a smartphone, and the rest of the devices are merely displays or shells? Sounds a bit aggressive but lets peel this back:

  • Apps: If you believe that the smartphone ecosystem and apps moves a lot faster than the rich client ecosystem, then that says that thin clients at some point will be able to run the same rich apps as a PC. Then the question becomes, "when".
  • OS/Dev Environment: iOS, Windows, and Android are all becoming modular, in that their goal is that you write once and deploy everywhere. Specifically, write once for a dev environment and deploy to a watch, phone, tablet, PC and TV or console.
  • Hardware: Fixed function blocks and programmable blocks on tablet and smartphone SOCs are taking over many of the laborious tasks general purpose CPUs once worked on. This is why many smartphones can display a beautiful 1080P video on an HDTV. This is true for video decode, video and photo cleanup, and natural user interfaces too. 3D graphics will continue to be an important subsystem in the SOC block.
  • Display: With WiDi, WiFi Direct, and WiFi AC on the mainstream horizon, there’s no reason to think that a user cannot beautifully display their apps from their 4" smartphone display to a 32" high resolution PC display. Today with my iPhone 4s airplay movieI can display 1024×768 via AirPlay mirroring with a little lag but that’s today via a router and WiFi network. I can connect today via hardwire and it looks really good. In the future, the image and fonts will scale resolutions to the display and the lag will disappear, meaning I won’t even need to physically dock my smartphone. It will all be done wirelessly.
  • Peripherals: Already today, depending on the OS, smartphones can accept keyboard, mouse and joystick via Bluetooth, WiFi or USB. The fact that an iPad cannot use a mouse is about marketing and not capability.

Smartphone Modularity a Sure Bet?

As in life, there are no sure things, but the smartphone and cloud ecosystem will be driving toward smartphone modularity to the point where they want you to forget about PCs. Apple, Microsoft, and Google are building scalable operating systems and development environments to support this. Why Microsoft? I believe they see that the future of the client is the smartphone and if they don’t win in smartphones, they could lose the future client. They can’t just abandon PCs today, so they are inching toward that with a scalable Metro-Desktop interface and dev environment. Metro for Windows 8 means for Metro apps not just for the PC, but for the tablet and the Windows smartphone. The big question is, if Microsoft sees the decline of the PC platform in favor of the smartphone, then why aren’t all the Windows PC OEMs seeing this too? One thing I am certain of- the PC industry cannot ignore the smartphone market or they won’t be in the client computing market in the long-term.

A Scenario Where Smartphones Take Down the PC

If you’ve done any long term strategic planning you know there are few absolutes but, very many scenarios. Tech history shows that even disruptive innovations take time to rollout and many scenarios existed that could have gone both ways. BlockBuster saw digital media coming and I will bet they had scenarios that showed varying levels of digital video acceptance showing what would happen to them if they didn’t lead levels of digital media leadership or lowest price. What if the publishers had stuck to their earlier guns and hastened digital rollouts? That could have given BlockBuster breathing room to develop more and they may still be around in their prior form.

There are other scenarios rolling out that are very interesting in that they could disrupt a giant, 500M unit market. That is, the scenario that has the smartphone “taking out” the personal computer.

I’d like to take a look at a few variables that could increase the likelihood of this happening. Remember, it’s not about absolutes, but about different scenarios and their chance of happening. Also, I’m not saying absolutely it will happen, but it is a viable scenario.

The New Personal
It all starts with the end user and making choices. If posed with the question, “if you had to choose between your phone of the PC, which one would you pick?” Sure, most want both, but making them choose makes them prioritize, and most would pick the phone. Why? One reason is that its so personal. People take it in the bed, bathroom, our pocket, on the dinner table. It knows where we are, what we’re doing, who we’re with, can communicate how we feel, etc. There are even reports that people would rather starve or refrain from sex rather than separate from their phone. Net-net, the phone is more personal and one variable that could, scenario speaking, accelerate the erosion and “take down” of the PC.

Good Enough Computing
Setting input and output aside for a second, the smartphone is pretty good, or good enough, for most email, web, social media, and light content creation. The web has actually “dumbed down” a but to make this possible and apps have helped almost as much Light content creation is writing email, editing photos, creating social media posts, and even taking notes. The big usage model exceptions to this are workstations and extreme PC gaming even though these are starting to be processed in the cloud. Most all else, scenario speaking, can be processed in the cloud.

Modular Designs
The iPhone 4s and the iPad 2 can already wirelessly mirror what is on the phone or tablet on the next best display. Most Android devices and even QNX can work with a full size wireless keyboard and mouse. Extrapolate that ahead three to five years with quad core general purpose processing, today’s console graphics capability, and even better wireless display technologies and it doesn’t seem, scenario speaking, that there won’t be a whole lot the user cannot do.

For “desktop” use, users will be connect to full size displays at high resolutions with full size keyboards, trackpads, and mice. Apple Siri, Microsoft Tellme and Google Voice Actions voice interfaces will be greatly enhanced in future iterations and can serve as the secondary input. Scenario-speaking, laptops could be wireless “shells” and leverage the processing power, graphics, memory, storage and wireless plans. The shells would cost a lot less than a full fledged laptop and have the convenience in that the content, apps, wireless plan is in one place.

One potential modular wild-card are flexible displays. While these have been demonstrated at every CES for over a decade, they appear to be getting very close to reality. While details are hard to come by, Samsung indicated that they will be shipping flexible displays in 2012. This could mean in phones by 2012 or shipped to OEM customers in 2012 for shipment in 2013. HP has been very active as well with their flexible display technology in alliance with ASU, the US Army, Du Pont, and E-Ink. HP is positioning their technology not only great for phones and watches, but also for larger POS displays, interactive advertising, and even on the sides of buildings. As it relates to smartphone modularity, think about “unfolding” a 10″ display from your 3″ device. That changes everything.

Potential Winners and Losers in Scenario
There are obvious winners and losers in this scenario. The big winners will be those who can monetize the smartphone or thin client and the cloud. Losers will be those who are stuck in the old model of computing, scenario speaking. If you’re one of those companies, I’d be rethinking your strategy.

Protectionism Rarely Works Over Time
Any scenario where well established and large losers exist, there will be protectionism. Over time, protecting something with such consumer benefit and such upside for other companies very rarely works. This is especially true for this scenario given the high levels of consumerism. Today, consumers have access to great info from the web and it’s amplified in the social media echo chamber. It’s hard to snow over consumers in any high value scenario.

Scenario Conclusion
The “smartphone kills the PC” scenario isn’t novel or new, but it is certainly one of the most important one of this decade. And certainly one of the most controversial as well given the 500M unit stakes with the winners and losers. How many of those will really be modular smartphones and how many will be PCs as we now it today?