There was a time just a few short years ago when the smartphone was considered a niche product and the convergence between voice, data and video merely wishful thinking. Industry watchers had seen it all – from the telecom world’s ISDN enterprise to the networking guys’ ATM pipe dream to cable mogul John Malone’s 500-channel multimedia universe. Pundits predicted that the smartphone concept would die faster than interactive TV.
Then came the iPhone. Now, the smartphone represents the latest golden age of consumer electronics and convergence at its best.
So what turned the dream of convergence into reality with smartphones? The answer lay partly in the history of the iPhone.
Steve Jobs approved the iPhone project only when Apple engineers assured him that ARM11-powered processors could handle the convergence of voice, data, music and video. EE Times chronicled the guessing games that Apple inflicted on the rest of the industry when it acquired PA Semiconductor in 2008 during the early days of the iPhone and the conception of the iPad.
Power was the nemesis of mobile devices and a key bottleneck to realizing handset convergence. Apple’s Power Semi acquisition was timely and played a key role in turning the iPhone’s newer models as well the iPad into marvels of successful integration. The iPad’s 10-hour battery life was widely credited to Power Semi’s chip designs that cut power use.
Indeed, it was basic building blocks like SoCs that turned the dream of convergence into a reality. The SoC methodology, which meshed multiple functions on a single chip, has been doing wonderful things for wireless and portable electronics devices since the early 2000s. SoC innovations not only enabled the convergence of voice, data and video, they also solved the power conundrum that haunted the seemingly ambitious smartphones. Apart from tighter integration of features like music and video, smaller footprints provided by SoCs led directly to a new architectural framework with integration at its core.
The semiconductor industry aimed to squeeze processor and memory devices along with much of radio functionality onto a single piece of silicon. Consequently, a core mobile phone chip promised improved battery life and almost five times the computing power of existing handsets. Combining flash memory with baseband and application processors removed the need for buffering between chips, leading to reduced power consumption and longer battery life.
One chip rather than three also meant a smaller footprint and hence more room inside the handset for other components. That advance lowered bill-of-materials cost of cell phones and freed up space for critical new technologies like GPS and Wi-Fi.
This was a major technological advance, one that remained largely hidden from the mainstream media. While the “trade press” has provided non-stop commentary about the software flash points like Android and Symbian, it also largely ignored the technological feat developed by chip makers in relative obscurity.
Although the conventional definition of a smartphone is closely tied to mobile OS platforms, when it comes to elevating user experience there is more to the heart and soul of smartphone than just mobile OSes. The fact that the ARM-vs.-Intel debate is now gaining the media’s attention is an acknowledgment of the crucial stakes for hardware design in consumer markets. And now that AMD is allying with ARM, the balance is likely to shift further toward the hardware side of the equation.
Mobile SoC suppliers like Qualcomm and Nvidia now hold considerable clout in driving future technology roadmaps for smartphone platforms like Google’s Android and Microsoft’s Windows Phone. These chip makers also have begun working more closely with third-party software houses to lay claim to the coming mobile revolution.
Meanwhile, multicore chip architectures under the SoC umbrella continue to evolve as processors, memories and other chips are integrated into new designs. The coming transition from 3G to LTE could provide the multicore movement with yet another opportunity to advance the strategic smartphone market.
–Former EE Times AsiaEditor in Chief Majeed Ahmad is the author of the new book, “Smartphone: Mobile Revolution at the Crossroads of Communications, Computing and Consumer Electronics”
It's just that I wanted to focus on SoCs in this particular article. Otherwise, in the book "Smartphone," for this integration marvel, I have mentioned SiP along with SoC. I have also listed memory stacks and dense PCBs as other critical factors.
With SiP, you need to wait for all two or three chips to arrive, but with SoC they effectively all arrive at once. But if the same company makes all two or three chips as different high-volume product lines already, that company usually prefers SiP.
An SoC can have many devices into one single piece. This is a key phase for miniaturization. Simply said, the embedded and portable technology couldn’t be real if it weren’t for the SoC’s.
You can’t have the software without the adequate hardware. Perhaps you could have the great ideas for the product, but if the hardware doesn’t have what it takes, the good ideas wouldn’t become reality.
One truth thing about software that makes it perhaps slightly over hardware is the versatility of software. In the same hardware you can re-program the chip to implement a whole plethora of ideas. And most of the times you can’t do that with hardware.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.