Large-Screen Compute Transformation Is Here

Mobile development never sleeps. Unlike some application areas where a technology takes root, grows and evolves quickly, and then plateaus when the products are mature, mobile is a different beast. It influences other areas of technology, bringing a new look and life to older applications.

Since 2009, each smart phone generation has delivered improved performance at the same or lower power consumption, more robust user experiences, and richer apps. Thanks to relentless ARM ecosystem innovation, each new premium phone migrated over time to different price points, opening new computing experiences to more (and sometimes previously smart phone-less) consumers.

Now we’re beginning a new chapter in mobile: the ecosystem fundamentally influences large-screen compute on tablets, convertibles, and other devices, including Android- and Windows-based systems. This is occurring, in part, because consumers now demand more of their smart phone “mobile” experience on those large-screen devices, including

• Always on and always connected capability
• Long battery life
• Thin, light-weight portable devices
• Rich graphics and media content
• Widespread apps availability.

Computex buzz
This year’s Computex shined a spotlight on these changes. Yes, the event was largely about high-performance computing, but the technical strides that companies in the ARM ecosystem have made in CPU and GPU designs in recent years mean that mobile innovation is more than ever influencing big-screen compute; and ARM’s presence at Computex was more pronounced than years past.

Anshel Sag, an analyst with the technology-research firm Moor Insights, said “ARM, a decidedly smartphone and IoT-focused company, was launching products that target PCs more than ever before.”

This was seen in major ARM announcements around the ARM Cortex-A75 and A55 CPUs, the first built on the company’s DynamIQ technology, and Mali-G72 GPU cores, both of which are designed with artificial intelligence in mind. The new cores target server chips and Chromebooks, as well as the upcoming Windows on ARM machines (more on this in a moment).

The announcements “solidly (put) ARM into the performance category of Intel’s mobile processors,” Sag said.

The technical achievements in these products are poised to propel mobile design further into what has traditionally been PC-design space. But what does that mean?

The Android effect
To understand the implications, think back just one year, to 2016, when Google introduced Android apps on Chromebooks.

With the availability of apps on Google Play, Chromebooks got one of the fastest growing, mobile-first app ecosystems in the world. Consumers are now beginning to engage with big-screen devices in the same productive ways they have enjoyed on their mobile devices.

For developers, the availability of Android apps on Chromebooks creates new markets. It will open new vistas for developers to push their apps into large-screen devices and to develop new sets of apps and use cases that leverage newer form factors and system performance.

Google Play applications on Chromebooks are available for multiple tablet and two-in-one form factors, however there are some significant challenges.

For non-ARM architectures to compete, they must translate the ARM native code into instruction forms within their architectural constraints. This comes at a cost in terms of performance, energy use, and user experience.

This impact on performance, battery, and other factors is illustrated in a significant study from Shrout Research, released recently as a whitepaper.

Shrout’s testing revolves around two Acer Chromebooks, the ARM-based R13 and the Intel-based R11. The ARM-based R13 is powered by a MediaTek M8173C SoC with dual Cortex-A72 cores and dual Cortex-A53 cores in a big.LITTLE. The Intel-based R11 runs the dual-core Intel Celeron N3060. Both systems include 4GB of system memory and eMMC storage typical of today’s Chromebooks. The Acer R13 has a 13.3-in 1080p display and the Acer R11 has an 11.6-in 1366×768 resolution screen.

The Shrout report concluded, “Of the two competing processor architectures for Chromebooks, the ARM-based processor tested in the Acer R13 exhibited significant and user-impacting benefits that include not only smoother animations and faster loading times, but fundamental compatibility advantages.”

You can read the full report here.

Another significant event in the large-screen compute in the past year has been the development of the OP Chromebook platform. ARM-powered OP devices, which run Android apps natively, are optimized for Chromebooks to improve performance, battery life, and functionality. Rockchip worked closely with ARM and Google to purpose-build the first OP SoC specifically for Chromebooks.

Enter Microsoft
Google and Android aren’t the only examples of this paradigm shift in large-screen compute. Qualcomm and Microsoft used their Computex presence to talk about their drive into connected PCs. They joined with ASUS, HP, and Lenovo in announcing technology based on the Qualcomm Snapdragon 835.

Sag wrote that the announcements represent “the fastest and possibly most high-performance application of ARM on Windows to date. So, even traditionally smartphone-first companies like Qualcomm were focused on delivering high-performance PCs from their own perspective.”

These directions are fundamentally rooted in the advancements that the mobile design ecosystem began to influence more than a decade ago. Today, one of its most significant impacts is in large-screen compute. Various names are associated with this, “connected PC,” “cellular PC,” among others. These designs exploit the cloud and traditionally smart-phone-focused connectivity and user experiences to revolutionize computing and deliver to the consumer what they want: access to their apps and data anytime, anywhere. The ARM ecosystem and a quarter-century of innovation lie at the heart of this transformation, delivering the performance, cost, and energy-efficiency requirements that consumers demand and, in the process, writing a new chapter in computing history.