MojoKid writes: Apple's new A7 SoC that sits at the heart of the iPad Air and iPhone 5S is a fast chip. Benchmarks and analysis have shown that it picks up its additional performance thanks to a mixture of architectural tweaks and, in some cases, its new 64-bit architecture. On average, the gains are split about 60/40 between the two areas, with more performance gains from the microarchitecture enhancements. But in one notable case — the popular 3DMark Ice Storm cross-platform benchmark — this hasn't been true at all. In 3DMark Ice Storm, the iPhone 5S is significantly faster in GPU workloads — almost 3x as fast in fact but its CPU performance is actually slightly slower than the A6, as measured in the Physics test. The iPad Air shows exactly the same performance issue, only its CPU is clocked faster than the iPhone 5/5S, and shows a small improvement as a result. Moving the code to 64-bit improved the A7's performance by about seven percent. The difference, it turns out, is tied to the open source Bullet physics library that 3DMark Ice Storm relies upon for testing CPU performance. While this doesn't dramatically change how the iPhone 5S ranks in 3DMark, it shows how the advantage of a big change (32-bit to 64-bit) can actually be much smaller than the impact of a low-level optimization that better matches how a CPU best performs a task.
As far as the laws of mathematics refer to reality, they are not
certain, and as far as they are certain, they do not refer to reality.
-- Albert Einstein