The mobile SoC market is highly dynamic and full of quite a few powerful solutions that have made their way into our flagship devices. We recently talked a bit about Nvidia’s Kepler-based Tegra K1, which will finally bring much of the company’s desktop graphics prowess into the mobile space.
Unfortunately, SoCs like the ARM Cortex-A15-based Tegra K1 (or its future 64-bit ARMv8-compatible Denver variant) are aimed squarely at the high-end. And rather than focusing on mid-range devices, most chip manufacturers instead re-purpose yesterday’s flagship architectures into today’s mid-range products. While this makes sense from a cost-cutting perspective for OEMs and chip manufacturers, it means that rather than enjoying newer and more power-friendly technologies, mid-range devices are stuck with comparatively second-rate products.
Now, ARM hopes to make a splash with its budget-friendly Cortex-A17 CPU core that is aimed squarely at mid-range devices. This is all while offering integer performance up to 60% higher than yesterday’s flagship Cortex-A9 CPU and incorporating today’s energy-efficient advancements and a 28nm die process. And on the topic of energy savings, the Cortex-A17 can be used in SOCs supporting ARM’s patented big.LITTLE technology for maximal power savings—a feature previously only seen on high-end chips.
Along with the performance gains and energy savings afforded by the Cortex-A17 CPU, mobile SoCs built around the A17 will also support rather advanced graphics by virtue of compatibility with ARM’s latest Mali-T720 GPU. But in an effort to make the CPU as versatile as possible, the A17 can also be used on existing AMBA3- and AMBA4-based systems and GPUs like the ARM Mali-450. Essentially, this means that you may end up seeing the Cortex-A17 CPU in a wide range of SoCs at varying price points and performance levels.
Now, let’s not get too excited just yet. ARM According to ARM’s own optimistic timetables, the earliest we can expect devices based on the Cortex-A17 to appear is sometime in 2015. This is because before the A17 can be used in shipping devices, it must first be incorporated into an SoC, and then that SoC must be incorporated into an actual device. Furthermore, it’s also up to the SoC manufacturers to harness the power and energy-saving technologies of the A17 effectively. That said, the prospect of a truly innovative mid-range chip (rather than a re-purposed former flagship) is legitimately exciting. We look forward to the creation of new and innovative mid-range products that chips like the Cortex-A17 will facilitate.