The thing is that the iPad Pro has the potential to blow all the PC tablets out of the water. It has the speed. It has the ram. It has the connection speeds between wifi and LTE. It just needs the apps. And I'm fairly sure that with the Pencil and the new keyboard, there has to be different development tools. I guess I see people thinking Windows or OSX apps when we talk about desktop apps. You can have full desktop apps on ARM. There are ARM implementations of Linux with full desktop apps. IT just takes a programmer sitting down with the new tools and deciding to do it.
So, I wonder about what "Desktop style apps" even means in this context myself. Are we talking about the mouse/keyboard UI? The more complete feature set that laptop/desktop versions of the apps have? The windowing system? Because iOS itself is a "desktop OS" with the windowing system (Window Server) replaced with Springboard. There isn't any magic in the code, really. The difference is in how you interact with the app, and what features you cut because it is
work to reimplement a big, old, legacy app on a platform that doesn't have the same UI concepts.
The first and third are a real long shot. They undermine the key identity of the device, and add complications that benefit specific workflows that are laptop/desktop centric.
The second one is to me the more interesting one, as that can be done. There are still limitations with how you can enable certain workflows on a touch-only device, and there are a lot of features that spawn dialogs on the desktop today which seem to be a no-no on iOS. Extensibility is a hurdle once you've sandboxed the OS like iOS, since any extensibility basically has to be blessed by the OS. I'd argue that neither Apple nor Microsoft has really done a good job of demonstrating what a fully-featured app looks like on the tablet. Nobody seems terribly interested in cracking that nut right now, oddly. Apple has the biggest incentive, and they are about as disinterested as anybody.
The "speed" is irrelevant when they're based on different architectures. As fast ARM still can't compete with an equivalent Intel chip.
That's what I don't get about people begging for an ARM Mac. Just because geek bench scores are comparable doesn't meant you can just switch them out. Even if the software still ran, there is years of interl architects optimizations and hyper threading technology.
Except, at the end of the day, when you look at a geek bench score, you are looking at some snippet of code, running on the CPU in some amount of time. It is measuring real work, even if it isn't the exact same kind of work users actually do. The code to test something like "how fast can the CPU compress/encrypt this block of data" is likely identical for both CPUs. So if that test shows similar
time, you are looking at similar CPU performance for that particular bit of code.
And honestly, HT tech is a bit of a trade-off. It helps in some situations (many threads, low CPU utilization), and hurts in others (high CPU utilization in apps that are properly parallel already). I really wouldn't call it something that lets Intel keep ahead of ARM. Where Intel beats the crap out of ARM is that it can just scale out to large TDPs and crush a CPU running at a fraction of the power draw through sheer force. But when the TDP is low, Intel's advantage evaporates.
It should also be pointed out that 32-bit ARM has been around since the 1980s. And compared favorably to Intel's 286 that was available, while also drawing less power. The main difference is that at some point, ARM stopped trying to compete with Intel directly, and stayed in the low-power space where Intel hasn't been a big player in until very recently. It isn't like Intel is the only one with decades of history and design work in their chip line. But when looking at the low-power regime, Intel is not the one with the track record in its favor.