In the weeks preceding the release of Apple’s new iPhones this year, rumors of the A8, the brains of iOS devices, speculated that Apple would finally give their A-series chips four cores and up the clock speeds to two gigahertz like Qualcomm and Nvidia had been doing for years. Apple, however, did not deliver in terms of these rumors.
Instead, the A8 was revealed to be still a 64-bit dual-core system-on-a-chip (SoC) with the most significant upgrade being the use of a 20 nanometer manufacturing process (Qualcomm’s chips still use the inferior 28 nanometer process) which significantly improved the A8’s performance and efficiency (by 25%-50%).
Pundits were disappointed, and misleading comparisons were made between the tech-specs of the iPhone 6 and the competing Android handsets outlining the low clock speed and number of cores of the A8.
Yet, the iPhone 5s, last year’s i-device, is still holds its own in benchmarks and real-world performance against the latest Android smartphones and phablets. What the A8 lacked in gigahertz and cores, it makes up in having a 64-bit design with ARMv8 instruction set (in contrast to the Qualcomm 801’s 32-bit design with ARMv7 instruction set) and 20 nm manufacturing process while still maintaining the power-efficient advantages of a dual-core processor. When Apple inevitably does make the jump to quad-core chips, its mobile products will truly have desktop-class CPUs while leaving Android products (none of which can use Apple’s proprietary A8) in the dust.
If Qualcomm and Nvidia wish to remain cutting-edge in the mobile processor industry, they must follow Apple’s lead and squeeze as much performance as possible out of their existing silicon–rather than relying upon superficial improvements–such as increased clock speeds and number of cores.
Written by Raymond Cao’17