Bay Trail was supposed to mark Intel's (INTC) ascension to the top of the mobile SoC heap. I believed it myself when I penned my articles "The ARM Singularity" and "Quantifying The Impact For Intel On Its Entry Into Mobile". But even at the time, there were a few doubts lingering, namely on the GPU side. As I said then:
Intel's possible failure
There is one ground where Intel might yet commit somewhat of a blunder. I'm talking about the GPU that's rumored to come with Silvermont. This is supposed to be based on Intel's HD4000 but with 1/4th the power, and thus at about the level where the iPad 4 is today. But unfortunately this market is fast-moving and where the iPad sits today is not where it will sit in one year's time. That's a possible weakness here.
I am afraid that those fears might have come to fruition. As Bay Trail starts getting its first benchmarks published, this is what we see (Source: Engadget.com, red highlight is mine):
This confirms what I said in my previous articles. That Intel's possible failure could be a function of the GPU performance. While there's still little data, Intel seems to have matched the competition on CPU performance and on energy consumption. But it still lags considerably on GPU performance.
Intel is only going to break into the mobile sector if it shows overwhelming superiority. With its lagging GPU such superiority is far from evident, so it's likely that once again Intel will see only limited success in the mobile market. At the very least, this development put off Intel's serious challenge for another year.
With the PC market still reeling from weak demand, Intel thus doesn't have any escape valve to compensate. And since estimates still discount a stable Intel, as seen below, a failure to enter the mobile segment in force thus presents downside (Source: Yahoo Finance, red highlight is mine)
The one risk where Intel could fail in its new generation of mobile SoCs seems to have manifested itself. Early benchmarks indicate Intel's GPU performance is not up to par. This should be enough to prevent Intel from overwhelming the ARM-based offerings already dominating the market.
Intel's serious challenge is thus delayed for yet another year. With the weakness in the PC market, INTC's equity is again a risky hold. The stock does have significant valuation support, but one cannot ignore that the catalyst for some upside seems to have failed at this point.