Intel Debuts Merrifield and Moorefield Designs at MWC
Performance
Performance on the two chips is expected to be competitive with other solutions, though Intel was markedly more coy about benchmark discussions than it was two years ago with Medfield. Test data was confined to the Intel-sponsored WebXPRT and MobileXPRT tests, with the older GFXBench's Egypt HD test thrown in for good measure.
Intel's decision to provide such limited data isn't very inspiring. Two years ago, when it unveiled Medfield, Intel was very open about where its products would place in industry-standard tests against competitive devices. Now, for whatever reason, it's eschewing that approach in favor of highly specific results. Though, one of Intel's engineers did post this video of a Merrifield-based reference platform in action, which seems to show the device performing well:
One additional point where Intel did give information, however, was the proposed speedup from adopting 64-bit Android applications.
Click to enlarge
Unfortunately, these estimations aren't actually based on 64-bit Android application performance, but on 32-bit vs. 64-bit runs of SPEC CPU2000.These results underscore the difficulty of making pronouncements about the evolution of the Android ecosystem, where app schedules and 64-bit launch rates are, by necessity, far more scattershot. It's not clear what the time table is for a unified 64-bit Android -- presumably we'll see the first 64-bit versions of the OS launch on devices later this year.
Looking To The Future:
Two years ago, we thought Intel had finally solved the problems that stood between it and effective competition in the smartphone industry. Intel's tablet plans, which called for a heavy focus on Windows 8 with first-generation Clover Trail, fell apart as consumers overwhelmingly skipped the high price and mediocre performance of these designs in favor of Apple and to a lesser extent, Android devices.
After Moorestown's 45nm disappointment and Medfield / Clover Trail's weak debut, we're a bit more wary of forecasting the third time around. We think Intel will gain tablet market share this year, due simply to the fact that it's putting a massive focus on marketing partnerships and working with the historic PC OEMs to do it. Dell, Lenovo, HP, and Asus have seen the writing on the wall for the conventional PC industry, and they're going to be eager to push adoption out to a wider customer base.
This time around, Android is an equal partner, not a secondary focus, tablet prices have come down to roughly half of where they were in 2012, and Intel is better positioned with new 22nm chips to launch into the market with a focus on classic partners rather than attempting to woo new OEMs. Still, it's an open question -- and a lot of eyes will be watching the chip giant to see how effectively it manages to sell into this space.
Intel's decision to provide such limited data isn't very inspiring. Two years ago, when it unveiled Medfield, Intel was very open about where its products would place in industry-standard tests against competitive devices. Now, for whatever reason, it's eschewing that approach in favor of highly specific results. Though, one of Intel's engineers did post this video of a Merrifield-based reference platform in action, which seems to show the device performing well:
|
Intel Merrifield Reference Platform |
;)
Click to enlarge
Unfortunately, these estimations aren't actually based on 64-bit Android application performance, but on 32-bit vs. 64-bit runs of SPEC CPU2000.These results underscore the difficulty of making pronouncements about the evolution of the Android ecosystem, where app schedules and 64-bit launch rates are, by necessity, far more scattershot. It's not clear what the time table is for a unified 64-bit Android -- presumably we'll see the first 64-bit versions of the OS launch on devices later this year.
Looking To The Future:
Two years ago, we thought Intel had finally solved the problems that stood between it and effective competition in the smartphone industry. Intel's tablet plans, which called for a heavy focus on Windows 8 with first-generation Clover Trail, fell apart as consumers overwhelmingly skipped the high price and mediocre performance of these designs in favor of Apple and to a lesser extent, Android devices.
After Moorestown's 45nm disappointment and Medfield / Clover Trail's weak debut, we're a bit more wary of forecasting the third time around. We think Intel will gain tablet market share this year, due simply to the fact that it's putting a massive focus on marketing partnerships and working with the historic PC OEMs to do it. Dell, Lenovo, HP, and Asus have seen the writing on the wall for the conventional PC industry, and they're going to be eager to push adoption out to a wider customer base.
This time around, Android is an equal partner, not a secondary focus, tablet prices have come down to roughly half of where they were in 2012, and Intel is better positioned with new 22nm chips to launch into the market with a focus on classic partners rather than attempting to woo new OEMs. Still, it's an open question -- and a lot of eyes will be watching the chip giant to see how effectively it manages to sell into this space.
