This started it all…
There was a recent flurry of articles about how Apple A4 is just “identical” to Samsung’s offering, based on the fact that both use a ARM Cortex A8 as the main processor. This is based on a report from UBMTechInsights which has published a photo of a sub-section of each chip. The photos shows two identical modules, namely the Cortex A8 main processor. Without seeing the whole die of both chips, this is meaningless. Most web reports are based on this comment from UBMTechInsights: “Initial investigation revealed that the ARM core used in the Apple A4 processor and the Samsung S5PC110A1 processor are identical”, which has been widely misunderstood. Read again: “that the ARM core [is identical]”
Apple’s A4 and the Samsung S5PC110A01 application processors are really the equivalent of a PC motherboard, with integrated graphics processors (GPU) and all. In this instance, the Cortex A8 would be the equivalent of an Intel processor. Saying that two PCs usethe sameCore i5 2.6Ghz processor hardly means anything about the overall capabilities of the computers. This is something that PC buyers are all too familiar with, and that makes the fortunes of PC Review sites.
The same goes for application processors: what GPUs do they have? Are there extra co-processors? How wide is the bus that links components to each other? Is there a hardware video-encode / decode? Are the blocks designed so that they can be shut down to save power? The A8 main processor might very well be only1 of 5 to 8 processors in there (Tegra 2 has nine).
The other blocksare outside of what’s visible in the above photo used by most reports. I’m not saying that the A4 is “better” than the S5PC110A01, or not. Each could be better in their own way. I’m just pointing out that we don’t have enough information to assert anything right now, including that they have similar capabilities.
…but a look at the whole picture showswild differences
However, a look at the *whole* picture reveals that both chips are indeed different. I’m not going to speculate on what each block is, but I suspect that Apple is more interested in saving power than it is in adding Mflops (a measure of raw speed) because iOS seems lean enough. Whichchip is better? It’s impossible to tell unlesswe benchmark them with a suite of software under the same OS. As you can guess, it’s not going to happen anytime soon.
In the end, it doesn’t matter much because you’re buying the device and the experience as a whole anyway.