Yeah sure, why not?. All the mainstream ARM architectures (ARM Cortex A77, Exynos 9820?, Apple A12) have surpassed Intel and AMD with respect to IPC. The only thing holding actual implementation of these architectures back in terms of desktop computing performance are the moderate clock frequencies as direct consequence of low TDP and hence low Voltage and high-density/high-Vth cell library mix.
That's oversimplifying things a great deal.
First, bi-endian mobile processor at low clocks and low TdP and small die has a far different development strategy from a little-endian desktop/prosumer 4/6/8/12/16 core processor that requires high base and boost clocks for heavy processing work, etc. ARM has been focusing on power efficiency. x86 has been focusing on processing power. So the two processes are fairly far diverged, and while the IPC are similar, good luck finding a current ARM processor that can overclock to 4 GHz to make those IPC similarities useful in the real world of desktop computing.
Second, there's a very good reason why multi-tasking is so limited even on the iPad - it's because engineering for multi-threaded performance is a lot different from the process for single-threaded performance. If it were so easy to toss an A12 into a laptop, Apple would have already done it. Instead, they
might be doing it (not with the A12, but with another ARM-based solution) in 2020, because it will require a reworking of macOS, Xcode, etc and they'll have to bake in backward compatibility like they were doing with Rosetta. It's not a small feat moving from x86 to ARM, though it's nice that they've had a lot of experience moving iOS toward simple multitasking on the iPad, so they can use that experience on the laptops. But it's going to be a lot of work. When I'm driving, my wife's iPhone with an A12 can play music, give me driving directions, and have a browser or Facebook open or Candy Crush. But the amount of background work that can be done while doing anything in the active forefront is really limited. On my home PC I can be transcoding Plex movies my kids are watching, have Youtube videos playing on one screen, and doing photo processing on the other screen with tutorials for photo editing open in the background in the browser, without a hitch. I am not sure an A12 can handle that at this point. But we don't know, because there's no real way to do these things on the A12.
Third, regarding performance on desktops, which are far more multitasking oriented, even if you look at AMD vs Intel multithreaded performance, you can see quite large differences in performance. Scaling up performance with cores is not easy. It's not like they can easily just ramp up the clock speeds and cores and it'll just work. So while the IPC are on-par with x86, the ability to increase clock speeds means they need to start redesigning their chip topology for power delivery to support those frequencies, take into consideration heat dissipation, develop a good architecture for inter-core communication which is going to be different for a 4-core or higher architectures than their A77x2 + A55x6 designs even. Their ISAs are entirely different, and optimizations for speed with x86 ISAs like SSE3 and so on mean x86 has a huge advantage for real-world tasks, and for ARM to reach on-par performance will take a lot more than simply gluing together cores and adding a beefy cooling solution.
Finally, there has to be software to run on the ARM processor. Windows on ARM is still a big issue. Even though chromium OS is compatible with ARM, ARM processors are still slow enough that few manufacturers are even attempting to use them in their Chrombooks or Chromeblets, and even when they do, they're clearly laggards in performance compared to similarly priced Intel offerings. Ubuntu has ARM-ready OS, but that's such a narrow use-case that I doubt it makes much of a difference. If/when Apple move macOS to ARM, that'll be a defining moment.
tl;dr: This is so much more complex than just increasing the TDP and clock frequencies.
Facebook
Twitter