Over the last few years, Apple has managed to set new records with its custom-designed A-series chips used in its iPhones and iPads. With the iPhone XS and iPhone XR this year, the company unveiled its 7nm A12 Bionic chip where it claimed a modest 15 percent performance improvement over the A11 Bionic from last year.
Given that Apple has always managed to introduce performance jumps of anywhere between 20-40 percent with its new chipsets, a 15 percent jump in CPU performance seemed modest. Well, turns out, Apple’s marketing team was extremely conservative in their claims.
AnandTech is out with its detailed review of the iPhone XS and they provide a fantastic insight into the A12 Bionic chip. Firstly, it is interesting to see that despite the A12 Bionic packing a staggering 6.9 billion transistors compared to 4 billion on the A11 Bionic, the former actually has a smaller die size — 83.27mm2 vs 87.66mm2. This is because of Apple making the switch to TSMC’s new 7nm fabrication process which has improved the overall transistor density.
In terms of frequencies, the two big ‘Vortex’ cores inside the A12 Bionic run at 2.5GHz compared to 2.38GHz on the A11 Bionic. However, only one big CPU core is able to reach this frequency, with the overall frequency dropping as more chips are required to run at full power. As for the four smaller ‘Tempest’ cores, they continue to run at the same speed as on the A11 Bionic: 1.6GHz. Apple has made some huge changes to the memory subsystem of its A12 Bionic which allows the big CPU cores to complete operations in considerably fewer number of cycles thereby bringing a major performance improvement.
The Vortex cores have only a 5% boost in frequency over the Monsoon cores, yet the absolute L2 memory latency has improved by 29% from ~11.5ns down to ~8.8ns. Meaning the new Vortex cores’ L2 cache now completes its operations in a significantly fewer number of cycles.
In terms of power efficiency, the A12 Bionic consumes 12 percent less power than the A11 Bionic at its peak performance in certain benchmarks. Overall though, the revamped memory subsystem consumes a lot of power when pushed hard which pushes the overall power consumption of the platform.
In the benchmarks where the performance gains were the largest – the aforementioned memory limited workloads – we saw power consumption rise quite significantly. So even though 7nm promised power gains, Apple’s opted to spend more energy than what the new process node has saved, so average power across the totality of SPECint2006 did go up from ~3.36W on the A11 to 3.64W on the A12.
Apple simply has no competitor in the mobile chipset department, with Qualcomm and Samsung offerings being considerably slower.
Overall the new A12 Vortex cores and the architectural improvements on the SoC’s memory subsystem give Apple’s new piece of silicon a much higher performance advantage than Apple’s marketing materials promote. The contrast to the best Android SoCs have to offer is extremely stark – both in terms of performance as well as in power efficiency. Apple’s SoCs have better energy efficiency than all recent Android SoCs while having a nearly 2x performance advantage. I wouldn’t be surprised that if we were to normalise for energy used, Apple would have a 3x performance efficiency lead.
For the GPU, when Apple claimed the 4-core GPU offered a 50 percent performance improvement, many thought the company had designed its own GPU from scratch. However, as per AnandTech, the GPU inside A12 — G11P — is very similar to that of the GPU found inside the A11. Apple has made huge performance improvements by adding memory compression, a surprising omission from its GPUs of the past.
The addition of memory compression along with an additional GPU core lead to the new 4-core GPU inside the A12 Bionic outperforming every other mobile GPU out there. This includes Qualcomm’s Adreno 630 GPU found inside on the Snapdragon 845 chipset which powers flagship Android devices like the OnePlus 6, LG G7, and more. In some cases, the performance improvement when compared to the iPhone X is well over 75 percent!
The improvements here are not just because of the microarchitectural improvements of the GPU, and having an extra core, all along with the new process node of the SoC, but also thanks to the new memory compression which will reduce power consumption of the external DRAM, something that can represent up to 20-30% of system power in bandwidth heavy 3D workloads. Saved power on the DRAM means more thermal envelope that can be used by the GPU and SoC, increasing performance.
Given the way Apple has been pushing the performance bar with its mobile chipsets over the last few years, it looks like the day is not far away when we will see the company unveil a MacBook powered by its own A-series chip instead of one from Intel.
The whole review of the iPhone XS from AnandTech makes for an interesting read so make sure to hit the source link below.