Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Apple’s 2½ year old iPhone 12 is 6% faster than the new Samsung Galaxy S23 Ultra

comparedial.com

Post reply on HN
Print view View original

501–509 of 509 posts

Re: Apple’s 2½ year old iPhone 12 is 6% faster than the new Samsung Galaxy S23 Ultra

#501

Post by goosedragons »

sidlls wrote: post #490
goosedragons wrote: post #462

Earlier quoted context omitted.

Android lets you easily disable notifications for some features of apps in settings. Apple is all or nothing and you have to hope that the app dev was kind enough to put granular notification settings in the app. It's also just better at stuff like in-notfication replies.

"Nobody" wants to spend 20 minutes configuring at that granularity: apple's is more than sufficient for 99.9999% of all use cases. All the supposedly superior Android features in this entire thread are things only a minute fraction of tinkerers actually care about.

Yes, everything you don't care about is some minute fraction of tinkerers. I bet you thought only some minute fraction of tinkerers wanted copy and paste, a notification shade, widgets, always on display etc.

Re: Apple’s 2½ year old iPhone 12 is 6% faster than the new Samsung Galaxy S23 Ultra

#502

Post by cubefox »

herpderperator wrote: post #51

> However, single-core performance is generally considered more important when it comes to overall speed for everyday usage, as most tasks are unable to scale efficiently across multiple cores. Thank you for acknowledging this. I think few people actually appreciate the need for better single-core performance. It's, in comparison, easy to just add more cores and use more power, but what is hard is making one core fas…

Counterargument: Single core performance is more important for most applications, but it is rather unimportant overall. Most phones have already more than enough performance for usual applications. What matters for performance is heavy applications, games, for which GPU and multi-core performance are important, not so much single core performance.

Moreover, if someone really is interested in the speed of normal applications, then it is more useful to directly measure things like start-up times of popular apps. Apple isn't ahead here, as far as I know, despite higher single-core performance.

Re: Apple’s 2½ year old iPhone 12 is 6% faster than the new Samsung Galaxy S23 Ultra

#503

Post by picture »

robertoandred wrote: post #366
picture wrote: post #13

Does benchmark results really matter that much for phones? I don't see myself compiling code, playing graphics intensive video games, multitasking with many engineering software running simultaneously, on a phone. Although Apple most definitely still leads on these factors, I think battery life, general latency/responsiveness of doing daily tasks, reliability, etc, are more important than some random coremark.

A faster benchmark means tasks get finished faster, which means improved latency/responsiveness and less battery used.

Late to reply (sorry) but this is an incorrect assumption. Benchmarks do not generalize to everything, especially loading apps from non volatile memory where users might notice latency the most. This is also usually a bottleneck by the memory related components instead of the core itself. Also, its not hard to just throw more brute clock speed at something, which will absolutely consume more energy overall than a slower but more efficient system.

Re: Apple’s 2½ year old iPhone 12 is 6% faster than the new Samsung Galaxy S23 Ultra

#504

Post by stolsvik »

kaba0 wrote: post #429
sliken wrote: post #220

Earlier quoted context omitted.

I believe it's still true. Apple's swift/obj c++ don't use garbage collection. Android default java/kitlin does, so android is inherently more memory intensive.

Ref counting is garbage collection. But it trades off better throughput for less memory overhead that is true.

You keep saying that, but it is not true.

Re: Apple’s 2½ year old iPhone 12 is 6% faster than the new Samsung Galaxy S23 Ultra

#505

Post by RulerOf »

asow92 wrote: post #204
RulerOf wrote: post #158

Earlier quoted context omitted.

This made me chuckle. I'm "still rocking" an iPhone 12 because I'm still paying it off. If it were dated and obsolete I'd be angry I threw away my money buying Apple.

Paid off the moment I bought it.

AT&T in the US offers "trade in rebates" on phones. You trade in a phone and they rebate your phone bill by $800 or $1000 over a 24 month period when you buy a new phone on an installment loan. It's a 0% loan so I'm not going to throw money away just to avoid paying less money over time.

Re: Apple’s 2½ year old iPhone 12 is 6% faster than the new Samsung Galaxy S23 Ultra

#506

Post by kaba0 »

stolsvik wrote: post #504
kaba0 wrote: post #429

Earlier quoted context omitted.

Ref counting is garbage collection. But it trades off better throughput for less memory overhead that is true.

You keep saying that, but it is not true.

Tracing GCs get to amortize their deletion costs over time - in theory a tracing GC with two equal spaces (thus at least 2X overhead) can just copy used data to the currently unused partition and switch over, later overwriting the former region. This combined with thread local allocation buffers where an allocation is only a pointer bump is a really great combo. There are many smart modifications, but this auto-defragments as well.

Now if you want general ref counting you have to use atomic counters, and those will trash your performance on modern machines beyond fixing. And then we didn’t even mention that big object graphs will have to be recursively freed, an overhead that can’t be amortized in this case. Oh and you do need a tracing step one way or another to free cycles.

Re: Apple’s 2½ year old iPhone 12 is 6% faster than the new Samsung Galaxy S23 Ultra

#507

Post by dzikimarian »

jacooper wrote: post #478
pjmlp wrote: post #475

Earlier quoted context omitted.

OP thinks it is. "It's at least $100 below market average and performance is great" => market average

He meant Market average as in average flagship prices. Pixel 6 pro and 7 pro prices are cheap compared to any other phone from the same tier.

This exactly. Regular Pixel 6 and 7 were also often mentioned as great value.

Re: Apple’s 2½ year old iPhone 12 is 6% faster than the new Samsung Galaxy S23 Ultra

#508

Post by jacooper »

dzikimarian wrote: post #507
jacooper wrote: post #478

Earlier quoted context omitted.

He meant Market average as in average flagship prices. Pixel 6 pro and 7 pro prices are cheap compared to any other phone from the same tier.

This exactly. Regular Pixel 6 and 7 were also often mentioned as great value.

Regular Pixel 6 has crazy value, you get an absolute steal for around 300$.

Re: Apple’s 2½ year old iPhone 12 is 6% faster than the new Samsung Galaxy S23 Ultra

#509

Post by stolsvik »

kaba0 wrote: post #506
stolsvik wrote: post #504

Earlier quoted context omitted.

You keep saying that, but it is not true.

Tracing GCs get to amortize their deletion costs over time - in theory a tracing GC with two equal spaces (thus at least 2X overhead) can just copy used data to the currently unused partition and switch over, later overwriting the former region. This combined with thread local allocation buffers where an allocation is only a pointer bump is a really great combo. There are many smart modifications, but this auto-defra…

I was referring to the “ref counting is garbage collection”. You do however seem to know your ways around a GC, so it seems we’ll just have to disagree on the ref counting == GC part. (And yes, GC rocks, but seriously bloats the memory requirements)
Post reply on HN