techconc

crowley said:

Wgkrueger said:

Who writes code like that? Having code that depends on the specifics of the CPU like the CleanMyMac X example is the very definition of bad coding practices. How embarrassing for them. 

I could I borrow your Coding Practices manual to see where it says that? And then maybe you could let the developers know what they should have done to fulfil the same function.

Moreover, it looks from the code snippet and explanation that the architecture is interpreting a request in a different way and returning a different result. How would the developer have know AppleSilicon would do that when they wrote the app years ago?

easy to be an armchair critic with the benefit of hindsight.

Actually, there are discussions like this on Apple's developer forum.  
https://developer.apple.com/forums/thread/12135

It's hard to tell what the developer was trying to do with that code snippet, but most would agree CFAbsoluteTimeGetCurrent() should be used.  I've used this with a benchmark application I was working on. 

Placed my order for Pacific Blue 512GB 12 Pro.   Wasn't trouble free experience though.  Tried both the Apple Store app and the web page.  Neither was available until about 6 minutes past the hour.  Eventually used the Apple Store app.  Transaction failed first time, before I even got to the payment section.  Had to try again.  Eventually got to the part where you pay.  Paid with ApplePay, but it took about 4 attempts before it went through.  I'm sure the store is getting pounded with orders, but still... not the best experience. 

cloudguy said:

Apple lacks their cloud prowess and infrastructure so they are trying to compete using hardware. 

No, that's not really true. iCloud has come a long way from Apple's humble cloud beginnings.  It now actually has a pretty good reputation.  Apple is not a generic cloud provider, but the services area where Apple competes, they do a pretty good job.  The real reason Apple does most of the AI on device is for privacy reasons.  Apple is ahead of there competitors with on device AI and that's only partly due to hardware advantages.  Apple has made a number of acquisitions for small AI companies like Xnor.ai which specialize in this field in order to gain a competitive advantage over Android, etc.  As privacy becomes more important, Apple's approach pays off in a big way.  In fact, what you see is companies like Google that are trying to catch up to what Apple is doing in terms of on device AI.

CheeseFreeze said:

This is more a modest evolution and not a ‘major increase’. Which is fine. 

I’m more interested in GPU performance and GPU features (and have that compared to AMD/Nvidia), as well as improvements to the Neural Engine side of things, because these factors will be way more telling of Apple’s advancements and strategy. 

And whether there will be a T3 chip following up the T2.

Who cares whether the CPU is just increasing in speed and efficiency, if the most useful and apparent factors are GPU and ML? 

As someone else noted already, the GPU compute scores are up about 70% in the A14.   The transistor count is up 39% which is pretty massive and it's clear this didn't go into the CPU so much.  Apple's focus seems to be exactly as you mention... with AI and machine learning.  This doesn't show up on Geekbench 5 as they don't have benchmarks to capture this.  There are industry benchmarks such as AIMark, etc. which do.  To that end, Apple doubled the performance of their Neural Engine and they now have 2nd generation matrix multiplication units used for ML.  The point being, much of the focus of the A14 has been on improving the AI/ML performance rather than traditional CPU where they already have a substantial lead.

As for a T3 chip?  No need.  The T2 chip was effectively a way to help bring Intel chips up to speed with some of the things Apple needs to do (Secure enclave, encryption, ISP, etc.).  The T2 is effectively based on the A10 technology and performance.  Apple Silicon Macs won't need a Tx chip because that's already built into their SoC.  I'm not expecting any significant new Intel based product introductions in the future. I'm guessing the last iMac update was the last hoorah for Intel on the Mac.

mykem said:

The single core alone is almost twice that of the Snapdragon 865 (1,583 vs 870).

The multi-core is a lot closer although you have to consider that the SD865 runs on 8 cores vs 6 on the Apple A14 (Apple has been using 6 cores on the regular A series chip since the A10- the “X” version for the iPad Pro uses 8 cores). For the multi-core the SD865’s GB5 score is around 3,280 vs 4,198 for the A12.

Outpacing Android based SoCs is nothing new for Apple.  I'd be surprised if that weren't the case.  What's interesting to me is that this is the first phone chip that has stronger singe core performance than anything from Intel.  Looking at Geekbench 5 Processor benchmark chart, the top rated chip for single core performance is currently the Intel Core i7-1165G7 which scores 1474.  For the first time, a phone CPU is now more than 7% faster than the fastest Intel desktop chip at single core performance.   We all knew this day would come, it's still a remarkable milestone when it happens.