iPhone 5S/5C thread

[shady28]

Crappy linpack implementation again, my Nexus 7 outscores that.

If they don't tell you what linpack app they are using there is no point in bothering with it because it's going to be Java.

So we can pick apart individual benchmarks, but frankly that can always be done, and its why these reviewers use more than one bench.

The 'bad benchmark' thought process falls apart when you have to also explain the 5s' 2x the sunspider performance of the G2, it's 80% faster on Mozilla Kracken, and nearly twice as fast on Google Octane.

Anyone who doesn't recognize that as a stunning game changer for the likes of google, Qualcomm, Nvidea, etc is deluding themselves. The only failure on Apples part here is putting it in a phone too small to really show off that accomplishment, but I would bet they'll fix that sooner rather than later.

 

[Eug]

Crappy linpack implementation again

I think we're starting to see a pattern here.

Android 4.1 on my Razr HD (which has near-stock Android) also suffers from mediocre implementation. It actually lags more in the UI than my iPhone 4 with iOS 7, which is shocking. I suspect a lot of that has to do with the lack of TRIM (which is present in Android 4.3, but again it's an issue of implementation. People can blame Motorola, but hey, Google owns Motorola.

 

[Accord99]

The 'bad benchmark' thought process falls apart when you have to also explain the 5s' 2x the sunspider performance of the G2, it's 80% faster on Mozilla Kracken, and nearly twice as fast on Google Octane.

Which are all basically the same thing, a test of Javascript performance.

There simply aren't any good cross-platform test of mobile CPUs in the mode of a SPEC that is used in consumer and server platforms and CPUs.

 

[Suspicious-Teach8788]

work escalations have killed the opportunity for me to go to the apple store in the morning... maybe i can call in to my meeting... ugh

 

[bearxor]

Cross your fingers on purchase online and pick up in store!

 

[Suspicious-Teach8788]

Cross your fingers on purchase online and pick up in store!

that will be a option? does the online store open midnight PDT tonight?

 

[shady28]

Which are all basically the same thing, a test of Javascript performance.

There simply aren't any good cross-platform test of mobile CPUs in the mode of a SPEC that is used in consumer and server platforms and CPUs.

So far you've decided to "discredit" linpack, Octane, Kracken, and Sunspider.

The pickings are getting thin.

How about the latest Geekbench 3?


Here's the Geekbench 3 score for the LG G2 / Snapdragon 800, a brand new Quad core "next generation" chip from Qualcomm :

Here's Geekbench 3 for the iPhone 5s vs other recent android flagships :

 

[Puddle Jumper]

I think we're starting to see a pattern here.

Yeah, sadly a lot of sites do a terrible job of comparing phone club performance, some due to ignorance and others due to intentionally stacking the deck.

If linpack shows a Snapdragon 800 getting less than 1gflop it's easy to see the test is broken, even the S4 Pro can manage that on native linpack.

Basically from what I have seen the A7 is close to Snapdragon 600 in performance, I haven't see enough to call it one way or the other especially considering the variations of S600 clock speeds available.

 

[Accord99]

So far you've decided to "discredit" linpack, Octane, Kracken, and Sunspider.

The pickings are getting thin.

That's because the pickings are thin for these reviews. Imagine if desktop CPUs were reviewed using only 4 javascript benchmarks.

How about the latest Geekbench 3?

Yay, a bunch of low-level algorithms, that are heavily dependent on how well they are optimized and compiled (how does Haswell only get 20.8 Gflop/s in DGEMM with 4 cores, a matrix multiply task, when it should be hitting 50+ Gflop/s with just one core), and whose scores can be greatly inflated by the presence of an AES accelerated component. At least with SPEC, people can compile it themselves for a true best-versus-best test.

Desktop CPU reviews have graduated years ago from using these things for anything more than FYI purposes.

 

[Ns1]

That's because the pickings are thin for these reviews. Imagine if desktop CPUs were reviewed using only 4 javascript benchmarks.


Yay, a bunch of low-level algorithms, that are heavily dependent on how well they are optimized and compiled (how does Haswell only get 20.8 Gflop/s in DGEMM with 4 cores, a matrix multiply task, when it should be hitting 50+ Gflop/s with just one core), and whose scores can be greatly inflated by the presence of an AES accelerated component. At least with SPEC, people can compile it themselves for a true best-versus-best test.

Desktop CPU reviews have graduated years ago from using these things for anything more than FYI purposes.

So basically, all benchmarks are worthless and we're left with real world performance metrics. Seems relevant:

Explain this to me.

I believe that is a HTC One, S4, iPhone 5 and some Lumia.

 

[Accord99]

So basically, all benchmarks are worthless and we're left with real world performance metrics. Seems relevant:

You mean a low-res animated GIF of 4 phones loading a game that you can't tell what it is unless you've played the game?

 

[Suspicious-Teach8788]

So basically, all benchmarks are worthless and we're left with real world performance metrics. Seems relevant:

Right and this is why UI lag has always been mentioned by me. I honestly don't care about how fast the CPU is theoretically, but if the experience is bad for users, then its a problem.

 

[Ns1]

You mean a low-res animated GIF of 4 phones loading a game that you can't tell what it is unless you've played the game?

YouTube video better?

I have a RAZR HD (1.5ghz DC Krait) vs an iPhone 5 on NFSMW and it's about a 20 second difference between the two to start playing.
http://www.youtube.com/watch?v=lq6DRr9MhLI

Can't wait for the 5s vs G2 vs S4 vs Note3 comparisons.

 

[QueBert]

So basically, all benchmarks are worthless and we're left with real world performance metrics. Seems relevant:

I call shens on that "test" that top left looks like Nokia's a WP device. Everyone knows there aren't any games on Windows Phone

YouTube video better?





Can't wait for the 5s vs G2 vs S4 vs Note3 comparisons.

Even better will be the comparisons next year when Phandroid dudes start making videos of the S5 & Note 4 vs the iPhone 5s. That's assuming the iPhone 5s still isn't faster, which it very well could be.

 

[MrX8503]

I have a RAZR HD (1.5ghz DC Krait) vs an iPhone 5 on NFSMW and it's about a 20 second difference between the two to start playing.
http://www.youtube.com/watch?v=lq6DRr9MhLI

Wow, that A6 surely does "suck" against Kait as PuddleJumper likes to put it. I'm sure the A7 will suck too as he's predicting.

 

[Eug]

My Krait device sucks terribly, but I blame Android, particularly Android 4.1.

 

[bearxor]

No pick up in store option

At this point I'm just holding my purchase. Will probably wait until I can stroll in to a Apple store and check it out.

 

[Suspicious-Teach8788]

No pick up in store option

At this point I'm just holding my purchase. Will probably wait until I can stroll in to a Apple store and check it out.

100 long at my local mall according to my buddy. There's never going to be an in store pickup option after the iPhone 4 launch issues. I wonder if I made the right choice by ordering online, but I unfortunately cannot take tomorrow off to work from the mall. Stupid coworkers scheduled 2 critical meetings in the morning. Ugh.

 

[st_7]

As usual, the guys over iFixit.com has given the just released Iphone 5s their treatment. They dissected it into parts & given a score of 6 out of 10 on repairability. Head over to this Link to know some interesting things about the internals of Iphone 5s & how the internal workmanship differs from Iphone 5.

 

[Imported]

Drove by the store in Palo Alto since its on the way to work.. must have been 400+ people in that line.

 

[Stuka87]

I ordered mine (Space Grey, 32GB) on the Apple store last night. Order was fully placed by 12:03am. Apple even let me keep my grandfathered in unlimited data plan (going back 4 years now to my 3GS, I have a 4S now). Ship time was 1-3 business days with free basic shipping (No option for faster shipping).

From what I have heard, there are some stores that only got a few dozen phones. And very few white or gold models.

 

[Oyeve]

Only $999 on the bay of e.

 

[bearxor]

Yeah, I didn't have any issues getting on the store this morning to try and purchase one. I just really wanted to pick it up today. I don't really know what I'm going to do at this point.

 

[Eug]

Only $999 on the bay of e.

One idiot in Toronto is trying to sell his 32 GB for $2000.

But maybe there will be an even bigger idiot that will pay it.

 

[st_7]

One idiot in Toronto is trying to sell his 32 GB for $2000.

But maybe there will be an even bigger idiot that will pay it.

Lol.. I agree, I think most idiotic purchases in the world usually happen on ebay.