8 ms·
Apple has done this twice already. 680x0 to PowerPC, then PowerPC to Intel. Adobe's come along on both of these.
by egypturnash 9y ago
Apple has done this twice already. 680x0 to PowerPC, then PowerPC to Intel. Adobe's come along on both of these.
- RcouF1uZ4gsC 9y agoThose changes were all to more powerful processors such that emulating the old architecture wouldn't be that bad. In this instance, Apple would likely be going to a less powerful processors and thus emulation at any reasonable level of performance wouldn't be that viable.
- schappim 9y agoThe processors wouldn't necessarily be less powerful. Case in point is the A11 outpacing the CoreM processors. The greater threat is the ios-ofcation of macOS. We've seen many companies pull their apps from the Mac App store due to the limitations of the sandbox.
- Bud 9y agoThat's not "likely" at all. Apple's CPUs are already faster in single-core benchmarks than the Intel competition, even though the Apple processors in question are designed for mobile devices and have much less power to work with. Apple would not be making this switch if they didn't think they could improve performance.
- drham 9y agoFaster, probably, but 2x+ Faster like in PowerPC > Intel very unlikely. Which is why there’s much less likely to be enough performance budget for effective emulation.
- saagarjha 9y agoPresumably their "desktop-grade" A-whatever chip will have a higher TDP and clock speed than their mobile counterparts.
- gascan 9y agoIf they could just turn up the TDP and clock speed on their chips and get 2x the performance of Intel's best chips that easily- they would already own the entire desktop market.
- mozumder 9y agoAnd some of us, with actual CPU design backgrounds, have been saying that for a while now. We've been asking why Apple doesn't already do it.
- samcat116 9y agoAs someone without a CPU design backround, would something like this scale pretty linearly with added power and thermal headroom? I assume there are limit that would have to be overcome, but what would an ARM chip in the conditions of an i7 look like?
- mozumder 9y agoIt's more of a logarithmic scale. The same architecture can generally scale to 10x over a few process generations.
- arbie 9y agoThis does not hold at 10nm and below.
- bzbarsky 9y agoTDP typically scales as somewhere between the cube and fourth power of the clock speed if you're pushing the envelope (in the sense of running at frequencies where further frequency increase also needs a voltage increase). So having 10x the thermal envelope means you can probably clock about twice as fast, all else being equal.
- BillinghamJ 9y ago
- mozumder 9y agoComputer architectures routinely see 3x performance jumps across different power budgets. This rule has held over decades. Clock speeds alone can probably increase by 30%. Caches and internal datapaths can double or more. Then you can start to add in more execution units or more expensive branch-prediciton or even new more power-hungry instructions. A 4 Watt Intel Pentium 4410Y Kaby Lake for mobile devices gets about 1800 on Geekbench, while a 115 Watt Intel Core i7-7700K Kaby Lake for desktops gets 5600. I'm just going to say it: the Apple laptop CPU is going to get Geekbench score... above 9000! And, yes, I do have a CPU design background.
- michaelmrose 9y agoSo artificial benchmarks already do a very poor job of capturing performance. The apple laptop cpu does not exist. If it did exist it would likely suffer a very substantial performance hit if forced to emulate x86 software. So you are going to speculate on the meaningless benchmark numbers of an imaginary cpu that will take a wholly unknown hit if everyone does not rewrite everything why?
- mozumder 9y agoArtificial benchmarks do a great job of capturing performance, since they're more controlled and eliminate unnecessary variables. Once you understand this, then you can understand how CPU designers work to predict future performance. CPU designers use artificial testbenches.
- michaelmrose 9y agoYou making up numbers doesn't appear to be a useful endeavor.
- sjwright 9y agoI suspect if Apple designs a desktop CPU, performant x86 emulation will be a key design criteria. I know very little about CPU design, but I imagine it would be possible to have hardware optimisations for x86 emulation just like we have today for video codecs. Or even further they could bake a "rosetta" into the chip's microcode and have their CPU natively support the x86 instruction set along with ARM or whatever they come up with.
- megy 9y agoDo you think they will use the same chip? In 2020?
- ajross 9y agoThat's not remotely true. There is some crossover, which is notable. But that means that the fastest iPhone is faster than some of the slower desktops. Intel's current single core leader is a coffee lake running at 4.7 GHz with like 12MB of core-speed cache. Be real.
- Bud 9y agoThe current A11 chip is faster single-core than the current MacBook Pro. The Intel chip you mention isn't in any Apple machine at all. (Except maybe the iMac Pro? Is it even in that?) So you be real. I don't think you've actually looked at the benchmarks. You should do so. I think you'll be shocked at just how much overlap there is, just like I was.
- LoSboccacc 9y agothat's because apple is fixed on choosing the intel parts with the least tdp power they can get away with. it's their choice to make the mbp slow, up and including the numerous thermal issues that force the cpu out of the nominal turbo mode for the longer workloads
- Bud 9y agoThose are valid points, but it's still shocking that Apple's own CPUs can outperform Intel laptop CPUs with far higher TDP.
- walshemj 9y agoThe 8700k which the previous poster mentioned is a hex core with 12 Threads!
- asdginionio 9y ago>I don't think you've actually looked at the benchmarks. What benchmarks? Geekbench? That's the only one I've ever seen where the fight is close, and Geekbench (like everything that boils performance down to one number) is nearly completely useless. If you see a two-watt phone CPU beating a 45-watt actively-cooled laptop CPU, you can either conclude that the phone is alien technology decades ahead of anything else on the market or that the benchmark is broken. Which is more likely? https://www.realworldtech.com/forum/?threadid=136526&curpostid=136666 https://www.realworldtech.com/forum/?threadid=136526&curpost...
- awalton 9y ago> Apple's CPUs are already faster in single-core benchmarks than the Intel competition Extraordinary claims require extraordinary evidence. "But Geekbench..." is not extraordinary evidence. Give me a board where I can do an Apple to apples comparison, booting both under the same revision of the same OS, and running industry standard benchmarks (read: plural), and then we can come to that conclusion.
- snowwrestler 9y agoNo one is stopping you from doing that experiment. > Extraordinary claims require extraordinary evidence. "Extraordinary" is in the eye of the beholder. We have some evidence; you're welcome to believe it or not, but unless you supply alternative evidence you're not really making an argument. You're just saying "it might not be true," which is something anyone can say at any time about anything.
- hcnews 9y agoNice, that was a good explanation for your stance. I am not supporting either, just liked the turnout. Obviously this isn't my usual kind of hn comment.
- ReverseCold 9y ago"booting both under the same revision of the same OS" It should be irrelevant in Apple's case anyway since they run their own OS. If their OS is faster for a certain task than some other OS then that's part of what you would define as "faster" for a user of the device. Right? Am I missing something?
- efficax 9y agoThere's no reason to think that the ARM chips they use won't be as powerful as their intel equivalents in a few years.
- saas_co_de 9y agoThis is true. The rate of progress in CPU performance has slowed to a crawl and will only get slower in the future. Because the pace of innovation of fabs has slowed Intel has also lost their fab lead and GF/Samsung/TSMC should all be very competitive with Intel over the next decade. Apple is probably looking at all these trends and concluding that they don't really need Intel.
- diogenescynic 9y agoAfter the Meltdown patch and Intel's exposed security flaws, Apple might still be better off with less powerful processors.
- mozumder 9y agoWhen Apple switches, their processors are likely going to be 2x faster than Intels. It will be a huge jump.
- CharlesW 9y ago> In this instance, Apple would likely be going to a less powerful processors… Arguably, the A11 Bionic is already faster than mainstream desktop CPUs. "The iPhone 8 even edged out the score from the 13-inch Apple MacBook Pro with a 7th-generation Core i5 processor. That notebook notched 9,213. Is Geekbench 4 really comparable from phone to desktop? According to the founder of Geekbench, John Poole, 'the short is answer is yes that the scores are comparable across platforms, so if an iPhone 8 scores higher than an i5, then the iPhone 8 is faster than the i5.'"[1] [1] https://www.tomsguide.com/us/iphone-8-benchmarks-fastest-phone,review-4676.html https://www.tomsguide.com/us/iphone-8-benchmarks-fastest-pho...
- lykr0n 9y agoGeekbench is a poor tool. Maybe for simple mathematical operations, but x86-64 has a larger instruction set than what the A11 supports
- Aloha 9y agoI dont think there is strong connection between a larger instruction set, and performance.
- Turing_Machine 9y agoYou are 100% correct.
- lykr0n 9y agoUsing a bad example, let's say x86-64 and A11 can both add numbers up to 3 digits. The a11 can do that faster than x86-64. But I want to add a 5 digit number. The x86-64 can do that with one instruction, but the A11 has to break that 5 digit number into multiple 3 digit numbers in order to process it. From my understanding, which might be wrong, Geekbench only tests what both can do. Comparing a pickup truck and a sports car you need to look at more than just how fast they can do. You can both move furniture, but the sports car will need to do a lot more to be able to what the pickup can.
- iagooar 9y agoARM less powerful? Wait and see what Apple can do with their own ARM-design if they target Desktops / Laptops. We might get to see 32 or even 64 cores in a regular MBP. Now tell me how THAT would be less powerful than the current Intel offering.
- saagarjha 9y agoApple has always gone down the path of fewer, more powerful cores instead of packing many, weaker ones. Apple's chips were dual core until A10.
- pilsetnieks 9y agoAnd it has let them eke out performance that absolutely leaves almost all Android phones in the dust, and all that with a battery that's 30-50% smaller than the rest (although that does bring its own problems, as we now know.) Imagine if they're not held back by the battery at all, or even just have to match the power consumption of the power-hungry x86 chips.
- imtringued 9y agoLet's be realistic. 64 A72 or A11 cores is a pipe dream. It's not going to happen without a massive nm process advantage. If we had a MBP with 64 wimpy a53 cores it would still only have 50% of the performance of a Intel i7 8700k or AMD Ryzen 1800X on paper. Of course that is faster than most notebooks but if you truly care about CPU performance then you usually wouldn't use a notebook in the first place.
- rahoulb 9y agoThere are also fat binaries - a recompile for apps that don't do anything too esoteric and you've got both formats in one bundle.
- clouddrover 9y agoMicrosoft is running 32-bit x86 applications on ARM today: https://channel9.msdn.com/Events/Build/2017/P4171 https://channel9.msdn.com/Events/Build/2017/P4171
- Corrado 9y agoI'm envisioning a Mac with multiple Ann chips in it. Why stop at 1 when you could put 8 or 14 Ann chips in it?
- armadsen 9y agoAs have Microsoft and other major companies. Unless you're writing large portions of the code in assembly language, porting to a new operating system can be greatly eased by the developer tools. In the past, Apple has provided a transitional emulation layer that meant that even architecture-specific code ran on their new machines.
- leejoramo 9y agoDon't forget Apple's NeXT heritage. NeXT supported and migrated between even more CPU platforms: 68k, x86, PA-RISC, and SPARC