With the introduction of the “New iPAD”, we now have the 2012 benchmark for the tablet market, including the offerings that will come from Amazon later in the year. As has been noted earlier, with each new mobile product iteration Apple unmoors itself from the PC foundations of Microsoft, Intel and even nVidia and AMD. At the unveiling of the new iPAD, Apple spent very little time extolling its wonderful, new A5X CPU and instead played up the experiences possible with their high definition Retina Display and high-speed 4G LTE communications. The number of ARM cores, the speed of the processors, the graphics engines are all of lesser concern than what the consumer experiences with the New iPAD connected to the Apple Ecosystem.
Back in the mid 1990s, when the world was trained to measure the value of their PC with the simple three letters: M-H-z and Intel was able to build an empire by delivering a new chip at a cadence of every 2 months backed by a full suite of synthetic benchmarks, there was the beginnings of an alternative vision of the future. The internet was just beginning to take off and just as significant, the bandwidth and wireless revolution that was well articulated by George Gilder in his book Telecosm pointed to a time in the future when the processor’s value would decrease relative to that of the communications infrastructure. Qualcomm was one company that Gilder tracked closely and looking back now almost 20 years later we can say it was his most prophetic selection.
Overlooked by nearly everyone in the 1990s was the importance and rise of the role of the graphics processor. Jen Hsun Huang’s vision of the future where the graphics processor becomes more important than the x86 processor was correct as can be seen in the increased die area dedicated to graphics within Intel’s Ivy Bridge and AMD’s APU offerings. After 4 cores, is there any value in an additional x86 CPU? The data says no. Intel’s transition from Sandy Bridge to Ivy Bridge is all about a vastly expanded graphics chip that for once challenges nVidia. As with processors, the graphics vendors measured their greatness with a full suite of benchmarks, which are now going to be cast into irrelevance in the new mobile markets with the New iPAD as the standard bearer. And soon to be followed by the iPhone 5 and a smaller $299 Retina display based iPAD in the fall. Did I mention the awesome Retina Display?
Because Apple is the leader in the mobile industry it is given great regard as to what is significant and supposedly true. When Apple’s Phil Schiller, Senior VP of Worldwide Marketing, displayed a graph that showed the new A5X with its quad core graphics is 4 times faster than the Tegra 3 without showing a benchmark, no one begged to question him. It is now written in hundreds of articles and plastered over the Apple web site. Personally, I am sure the Tegra 3 stacks up well to the A5X.
On the day after the launch, the technical web site Tom’s Hardware asked nVidia if it was true that the A5X was 4 times faster than Tegra 3. The response was telling in that nVidia wasn’t willing to challenge Apple’s claim on benchmarks without hardware to test. Should nVidia find out in the coming weeks that the Tegra 3 is indeed on a par or better than the A5X, I doubt that they will make it public. To do so would jeopardize their attempts to win the next generation MacBook Pro graphics sockets. And so they face a dilemma of losing their benchmark marketing tools in order to remain a supplier. Apple has effectively de-positioned nVidia’s Tegra 3 that is slated to be in a number of Android based tablets and smartphones this year. The Tegra 3 was smothered in the cradle. Qualcomm and Intel will be the benefactors of this deliberate Apple Branding slight of hand.
As I wrap up this blog posting, TI has just reported that sales will come in less than previous guidance based on slowing OMAP sales. In addition there are reports that Broadcom is challenging TI for the sockets for the next Amazon Kindle. Combine this with Intel’s aggressive push into ZTE, Lava, Orange, Lenovo and Motorola and one gets a picture of a fast commoditizing application processor and graphics market. The one chip inside the mobile smartphone and tablet that demonstrates high value is the baseband chip, which is being dominated by Qualcomm as they start to ramp 4G. Intel knows this and has a massive R&D effort underway to catch them in time for the launch of Haswell based ultrabook platforms in 2013.
Recent remarks by Intel executives appear to support the position that Intel would gladly torch the market with which the Atom Medfield processor is designed to serve in order to eliminate nVidia, Broadcom, TI, Marvell and AMD as viable competitors and more importantly to increase the data center footprint that spits out an unbelievable 50% operating margins. Qualcomm and Intel share a common goal of ramping 4G LTE into mainstream price points so that the Mobile Bandwidth Tsunami causes profits to rain down on the two ends of the wireless network.
The Era of Benchmarks is Coming to an End or I should say the new mobile market has tossed aside the benchmarks with which the PC market operated under for the last 20 years. I expect the Ultrabook PCs due out this year with their Retina based touch screeens and extended battery life to follow the pattern set by smartphones and tablets. The semiconductor playing field continues to experience rapid and dramatic change.
Full Disclosure: I am Long AAPL, QCOM, INTC, and ALTR