When a new Apple-branded product launches, it usually doesn’t take long before it gets disassembled and dissected. The iPhone 6 has already been featured in its own teardown, and so has the larger iPhone 6 Plus. Now, though, it’s time for the A8 processor to go under the knife.
According to a report published by AnandTech, alongside the teardown performed by Chipworks, the A8 SoC has been disassembled and given an in-depth investigation to what makes it tick, and the features it holds. As noted by AnandTech, the A8 SoC does not include the Imagination Power GX6650, but instead boasts 4-core GX6450. This isn’t that big of a surprise, though, as the GX6450 is a direct successor to the G6430, which was what Apple utilized within the A7 processor. Apple even managed to save some space with the the newer GX6450, as the 20nm process measures in at 19.1mm2, versus the 22.1mm2 measurement of the A7 chipset.
One other key detail found by Chipworks is the SRAM cache memory. The teardown revealed that the 4MB block serves the same purpose as it did on the A7, in servicing the CPU and GPU. AnandTech notes that while the SRAM takes up a “fair bit of space,” there is an obvious value to being able to serve larger memory requests without having to go off-die to make it happen:
“This 4MB of SRAM ends up being quite big despite the shrink from 28nm to 20nm, and while at first glance it seems like it should be larger than 4MB given the relative size, in practice what has happened is that the individual SRAM cells have not shrunk by a full 50%. Chipworks estimates the cell size to now be about 0.08µm2, versus 0.12µm2 on A7, which is closer to a 33% shrink that a 50% shrink. As a result the SRAM cache still takes up a fair bit of space, but the value of being able to serve larger memory requests without having to go off-die continues to be immense.“
One final detail includes the affirmation that Apple’s A8 was indeed fabled on TSMC’s 20nm process. As noted, this makes Apple’s current-generation SoC one of the first to be fabricated on this specific process. The result means that Apple was able to shave off about 15 percent of the die size, while managing to include even more features.
You can check out the full report through the source links below.[vai AnandTech; Chipworks]