4 ms·

There are several aspect to the perf hypotheses you make, and I don't see why Apple could not design its own CPU with excellent perfs. Current Apple best CPUs

by wilun 9y ago

There are several aspect to the perf hypotheses you make, and I don't see why Apple could not design its own CPU with excellent perfs. Current Apple best CPUs are already performing very well compared to some Intel CPU. Not the fastest Intel CPU, obviously, but a) Apple Arm implementation is already above all the Arm competitors b) we are talking about an Arm implementation optimized for low energy consumption and other components and functions essential to a high end mobile phone -- and even then it starts to catch up chips that are dissipating more and integrate far less features c) the hypothetical Apple latop/desktop cpu won't be coming today. From a theoretical pov I see nothing that prevent Apple to achieve what they want. Especially with the kind of money they have, and the kind of team they have to already design chips. Oh btw, the A11 has more transistors than a 4core Skylake. That does not means much, except that we are dealing with the similar kind of projects, and that in that kind of project, the perf and thermal budget you have is not something that suddenly makes the people working on it go: oh my god, but we don't know how to make that, we are only good for mobile phone SoC. I suspect some of Apple employees working on their chips even worked at Intel before... Another problem with what you say is that the workloads you consider are especially well suited for accelerations, and it is even simpler to compete in this area compared to old school CPU power. You just need bulk compute power, that can come in the form of, at least: vector instructions; GPU; DSP; fixed function acceleration units. Those areas are not especially dominated by Intel. Some last points are that: top CPU speed are (mostly) not growing anymore. That lets others catch up Intel in the few area that they did not yet -- others also are catching up on the process side. The result of that equation is that I really don't see why Apple could not manage to create high performance CPU to the point of not needing Intel anymore. Maybe at first they would limit to their laptops, and switch the Mac Pro last, but I'm not even sure of that.

imtringued 9y ago
>Some last points are that: top CPU speed are (mostly) not growing anymore. That lets others catch up Intel in the few area that they did not yet -- others also are catching up on the process side. Meanwhile in reality the top end non apple ARM SoCs only have the performance of a single highend intel core. http://browser.geekbench.com/android-benchmarks/ http://browser.geekbench.com/android-benchmarks/ vs http://browser.geekbench.com/v4/cpu/7779910 http://browser.geekbench.com/v4/cpu/7779910
- wilun 9y ago
  Yes but Apple has an edge compared to other ARM (and that's what we've been discussing about in the first place) and comparing the CPU speed of an energy optimized phone SoC to an high end desktop core is absolutely not representative of what you can expect from a team tasked with creating a CPU core dedicated to performance in the first place, with less power constraints. (And that's also why the work Apple has already done on their own CPU is so impressive, btw.) Otherwise I could as well take the perf of an IBM z14 mainframe and compare it to an existing smartphone SoC, and declare that Arm is doomed because too slow. Note than even increasing the clock alone of those ARM cores would increase the perf greatly -- it might not be possible without modifying them in various degree right now (if you don't target high freq to begin with, you can afford to work less on the length of some critical data path), but given today processes it is very probably not extremely hard to tune the design if needed.

by wilun 9y ago

imtringued 9y ago

>Some last points are that: top CPU speed are (mostly) not growing anymore. That lets others catch up Intel in the few area that they did not yet -- others also are catching up on the process side. Meanwhile in reality the top end non apple ARM SoCs only have the performance of a single highend intel core. http://browser.geekbench.com/android-benchmarks/ http://browser.geekbench.com/android-benchmarks/ vs http://browser.geekbench.com/v4/cpu/7779910 http://browser.geekbench.com/v4/cpu/7779910

wilun 9y ago
Yes but Apple has an edge compared to other ARM (and that's what we've been discussing about in the first place) and comparing the CPU speed of an energy optimized phone SoC to an high end desktop core is absolutely not representative of what you can expect from a team tasked with creating a CPU core dedicated to performance in the first place, with less power constraints. (And that's also why the work Apple has already done on their own CPU is so impressive, btw.) Otherwise I could as well take the perf of an IBM z14 mainframe and compare it to an existing smartphone SoC, and declare that Arm is doomed because too slow. Note than even increasing the clock alone of those ARM cores would increase the perf greatly -- it might not be possible without modifying them in various degree right now (if you don't target high freq to begin with, you can afford to work less on the length of some critical data path), but given today processes it is very probably not extremely hard to tune the design if needed.