[ 3 / biz / cgl / ck / diy / fa / g / ic / jp / lit / sci / tg / vr / vt ] [ index / top / reports / report a bug ] [ 4plebs / archived.moe / rbt ]

Due to resource constraints, /g/ and /tg/ will no longer be archived or available. Other archivers continue to archive these boards.Become a Patron!

/g/ - Technology

View post   

[ Toggle deleted replies ]
File: 32 KB, 527x329, BTFO.png [View same] [iqdb] [saucenao] [google] [report]
50442091 No.50442091 [Reply] [Original] [archived.moe] [rbt]

Why haven't you upgraded to the latest and greatest /g/?

>Dual-core iPhone 6s obliterates Galaxy S6, Note 5 and other top Android phones in performance tests


>“The Only Thing That’s Changed Is Everything”


Poorfags need not apply.

>> No.50442163

No file system. What's the fucking point of having 128GB if I can't even view my video files.

>> No.50442211

>single threaded performance

Congrats do you want a cookie for living in the past still?

>> No.50442283


>the iPhones 6S still have only two cores, but single-core performance is a better measure for the sort of things you can feel while using a device. Apple is literally years ahead of the industry. Even if you don’t agree that single-core performance is the more meaningful benchmark, you can’t deny that iPhone benchmarks don’t look anything like the benchmarks for high-end Android devices.

Your multi-core score is just dick-waving. Single-core efficiency gives better performance. LEL LAGDROID AMIRIGHT

>> No.50442301


* https://daringfireball.net/2015/09/the_iphones_6s

>> No.50442345

>gets shit on in multi
>shut up multi doesn't matter at all its all about the tests we do good at

Spoken like a true idiot from the 90s

>> No.50442414
File: 497 KB, 1887x1024, Android_BFTO.jpg [View same] [iqdb] [saucenao] [google] [report]

Use this image instead. Turns out only 8-core android phones can surpass iPhone 6s multi-core performance, though there aren't that many. iPhone 6s beats all androids in single-core benchmarks though.

>> No.50442441
File: 19 KB, 390x470, laugh crying.png [View same] [iqdb] [saucenao] [google] [report]

Multi doesn't mean shit if there's no optimization to use it.

How did AMD's MOAR CORES!!1 work out poorfag? Oh that's right, so futuristic to have more cores out of your ass and still lose to a more efficient design with less cores.

>> No.50442493

geekbench is not comparable across platforms you ignorant shit. Unless you actually believe the new iphone outperforms macbooks (pro tip, it doesn't)

>> No.50442522

Please stop making shit up, we have proof 8 cores IS actually a useful thing to have for android, and further we can see improvements can be had from even 10 or more core designs.

Read something informative, you'll look less stupid next time you try and make shit up.

>> No.50442530

holy shit the shills are in full force today
> falling for the botnet
> being a normie

>> No.50442539

Yeah you're right, no one renders shit or rips DVDs from disc anymore....they only play poorly optimized flash games and rub thier iphones on thier manginas

>> No.50442565


>Cross-platform processor benchmark
>Available for Windows, Mac, Linux, Android, and iOS

>> No.50442605
File: 25 KB, 650x425, Jim-Keller-AMD.jpg [View same] [iqdb] [saucenao] [google] [report]

Just wait and see what more cores will do.

>> No.50442614


Made no shit up at all butt-blasted android user. No doubt more cores are useful if systems and apps are made to take advantage of resources available.

You can't not say that single-core is still king when the system or apps do not take advantage of multi-core

>> No.50442636


Yeah just let me rip a DVD and render some studio quality video on my android. LEL

>> No.50442658

Wtf it's not a tri core?

>> No.50442661

Just because it runs on multiple platforms does NOT mean the results can be compared across platform you fucking retard.

I mean, do you really think a phone single core performance is the same as an intel desktop chip? The intel i5-3330, a 3.2Ghz quad core desktop CPU, scores 2400 on geekbench for single threaded performance. That's exactly what the new iphone gets as well. And you HONESTLY think they are comparable... are you fucking stupid or retarded?

>> No.50442665


>Geekbench 3 scores are calibrated against a baseline score of 2500 (which is the score of an Intel Core i5-2520M @ 2.50 GHz). Higher scores are better, with double the score indicating double the performance.

>> No.50442681


See >>50442665


>> No.50442743

... you're not helping your case at all m8, now you're saying an intel i5 CPU is just as powerful as the iphone, which again is just wrong.

Unless apple has managed to break the laws of fucking physics in the last year when I wasn't paying attention.

Do a computation like super pi, have the CPUs calculate pi out to 32M decimal places. The intel laptop CPU will easily win, every single time. Despite that on geekbench the new iphone has almost identical single core performance. The intel CPU has way more ACTUAL single core performance.

Geek bench is a shitty tool to compare outside of the OS. Compare android to android or iphone to iphone, or windows computer to windows computer, but you can't compare them to each other and expect it to represent anything meaningful.

>> No.50442998


>an intel i5 CPU is just as powerful as the iphone

No you're being vague as hell.

>Geekbench 3 scores are calibrated against a baseline score of 2500 (which is the score of an Intel Core i5-2520M @ 2.50 GHz). Higher scores are better, with double the score indicating double the performance.
>calibrated against a baseline score of 2500 (which is the score of an Intel Core i5-2520M @ 2.50 GHz)
>Intel Core i5-2520M
>mobile cpu
>not desktop

Geekbench is cross platform. Don't get confused with 'multi platform'. Scores are base-lined with that mobile processor, which the latest and greatest iPhone now surpasses.

>Geek bench is a shitty tool to compare outside of the OS. Compare android to android or iphone to iphone, or windows computer to windows computer, but you can't compare them to each other and expect it to represent anything meaningful.

Says you? Or the benchmark tool which concisely measures across platforms?

>> No.50443086

>>calibrated against a baseline score of 2500 (which is the score of an Intel Core i5-2520M @ 2.50 GHz)
This is my point, an i5-2520M scores 2500 single core.
The new iphone 6s scores 2400 single core.

So that means the new iphone 6s MUST be almost as powerful as a laptop CPU. Which is just factually wrong, it isn't anywhere near as powerful as that CPU.

The intel CPU has a TDP of 35 watts

the A9 (if its similar to previous SoCs from apple) will have a TDP around 3-4 watts.

Just do the math, even with apple being on a newer process (16nm vs 32nm)

Also, how the fuck are you gonna directly compare ARM cores to x86 cores? you fucking mong. Without the source code for all you know the benchmark is optimized for ARM cores and totally shit for x86

>> No.50443189

Since there are different architectures involved, you are just comparing a very specific piece of code and nothing else. Obviously a cellphone CPU isn't as powerful as a laptop one in general tasks. The geekbench score just means that THAT SPECIFIC PIECE OF CODE runs at a similar speed on both processors and nothing else.

>> No.50443225


Who went of on a tangent and was comparing the iPhone6s to more powerful desktop class cpus - oh that's right, you.

From the OP, the comparison was against Macbooks:

>Combing through Geekbench 3 results, the 1.3GHz MacBook’s scores compare most directly to Apple’s 1.4GHz Macs, such as the entry-level 21.5″ iMac and early 2014 entry-level MacBook Air. The latter model achieved Single- and Multi-Core scores in the 2400/4700 range for 32-Bit tests, and 2700/5300 for 64-Bit tests.

>Speaking of the Mac and performance, again looking at single-core performance, Geekbench results put the iPhone 6S at roughly the same score as a 2012 or 2013 MacBook Air.

Now, what manufacturer supplies laptop cpus to Apple.. hmm, couldn't be intel, eh? As for the benchmark being optimized for either ARM or x86 - puh-lease tripple-nigger. It wouldn't be a very good benchmark then would it, princess?

Android still BTFO!

>> No.50443253


So what you're saying is that by having a piece of code that only tackles certain criteria (integer, floating-point performance) is better than taking advantage of specific cpu exclusive instruction sets! Genius! Why haven't those Geekbench guys thought of that!

>> No.50443262

>Apple is literally
I thought that meme was only used on 4chan?

>> No.50443269

jesus you really are this stupid.

No wonder iphone faggots stick with iphone, they have no actual knowledge of how hardware fucking works.

You're simply not understanding the core idea here. You're comparing two totally different architectures with a benchmark that doesn't let you see what its actually running to get the magic numbers at the end. And you HONESTLY believe that just because this silly magic number matches or beats a laptop from 2012-2013 it is actually more powerful than those laptops...? I mean seriously is THAT what you fucking think? Because if so you just have no grasp of how computer CPUs work.

>> No.50443270

>Applel shills at it again

>> No.50443280

I can't wait to see the real ARM shitwrecking when keller's k10 is out.

>> No.50443300


>> No.50443328


Why bother comparing them in the first place using Geekbench then? Are you saying that all people that run benchmarks comparing between architectures are being morons and that Geekbench should give a disclaimer saying, "Hey, don't compare calibrated, base-lined scores with any different architecture whatsoever!"? Imagine that, all this time tech sites were getting it so wrong.

>> No.50443331
File: 2 KB, 73x46, 1441149709845.jpg [View same] [iqdb] [saucenao] [google] [report]

How does an a72 core preform again the latest apple "cyclone"

>> No.50443339

>How did AMD's MOAR CORES!!1 work out poorfag?

Except AMDs underlying architecture was shit. Unless you are actually implying Android phones use AMD chips.

Oh wait /g/ is now a /reddit/ normie haven that has no clue about cpus to begin with and think facebook and flash games is all that a phone is for.

>> No.50443363

>all this time tech sites were getting it so wrong
except no respectable tech site uses geekbench to compare across platform. Literally no respectable tech site does. Point me to a single fucking one, I dare you.

>> No.50443364


>Geekbench Score The Geekbench score is the weighted arithmetic mean of the three section scores. The Geekbench score provides a way to quickly compare performance across different computers and different platforms without getting bogged down in details

Oh, look at that.

>> No.50443387


>Except AMDs underlying architecture was shit

Exactly my point, dilbert.

>> No.50443404




What do you class as a respectable site?

>> No.50443406

>without getting bogged down in details
Because the details are what actually matter for telling you how powerful the CPUs are.

An intel i7 4770k scores around 4,000 single core. So it must only be about 80% faster than the new iphone right?

Pssh, a $300 95W desktop CPU isn't even TWICE as fast as 3-4w $20 SoC from apple?

Seriously, just fuck off if you don't understand how the fuck hardware actually works.

>> No.50443417

He didn't say anything about desktop CPUs.

Also I don't think you know anything about software optimization or CPU architecture.

>> No.50443423


I understand completely how it works. Did you even read the link I referenced you bogan?

>> No.50443426

>weighted mean

what are the weights?

And why should I believe GeekBench is anywhere near accurate when SPEC 2006 has 29 benchmarks in order to actually show off the power of CPUs?

>> No.50443427

>no respectable tech site uses geekbench to compare across platform
>gives 2 examples of sites comparing ONLY between apple devices

ok? Thanks for proving my point m8

>> No.50443442


My dad is bigger than your (trans) dad

>> No.50443456


You asked for tech sites that used Geekbench. At least Anandtech is one of the most respected. I win.

>> No.50443464

Except he's right. I read your link and it clearly says better score -> more performance where double score directly correlates to double the performance.

>> No.50443484

>A9X faster than AMD Turdozer at a fraction of power consumption


>> No.50443498



>> No.50443514

>You asked for tech sites that used Geekbench.
No I didn't fuck head, i specifically said

>no respectable tech site uses geekbench to compare across platform. Literally no respectable tech site does. Point me to a single fucking one

Like I fucking said, point me to one that compares across platform. I know you can use geekbench to compare between 2 iphones, or other android devices between other android devices.

I specifically said show me an example where they are comparing between platforms. Pro tip, they fucking dont do it because it doesn't give any meaningful results.

>> No.50443515


He asked for ones that compare CROSS PLATFORM you fucking iShill double nigger faggot

>> No.50443598

Apple hardware is fucking nice. Too bad it runs a Turd sandwich for an operating system. If they made ios a bit more open for people to tinker with, I would never purchase another android, ever.

Just give me a low level access to kernel and user land. It doesn't have to be by default, but give me the option to hook it up to my iTunes or whatever and select an "unlock" button that turns off signed code verification and gives me root level access. I would be a customer for life.

>> No.50443623


Like this? http://www.notebookcheck.net/SoC-Shootout-x86-vs-ARM.99496.0.html

Now you'll comeback with, "pah, it's not a REAL tech site!"

Here's another:

>The folks up north at Primate Labs have some benchmarks of a PowerMac G5 and a Core 2 Duo iMac running under Tiger and Leopard. They used their cross-platform benchmark Geekbench for the tests.


>> No.50443655

Yes, I read your link and it makes no mention of what the actual weights are, only that they are weighted. Nor does it make it any better of a measurement compared to SPEC 2006 which is an industry standard for comparing CPUs.

>> No.50443699

>Like this? http://www.notebookcheck.net/SoC-Shootout-x86-vs-ARM.99496.0.html
That kinda proves my point

Snapdragon 600 overall score in geekbench of 3185

AMD A4-5000 overall score 3051

So by that logic, the SD600 should score slightly better than the A4-5000 overall right?

WRONG. Just look at the rest of the benchmarks, the A4-5000 CRUSHES the SD600 in pretty much everything, sunspider, octane, go down the list and the A4-5000 is generally TWICE as fast..

B-b-b-uh muh geekbench

yeah fuck off

>> No.50443730

>trying this fucking hard to prove that guy wrong
Lmao wtf is wrong with you? Without looking it up, what are those numbers a measurement of? Exactly, stop talking out of your ass haha

>> No.50443743


looks like someone got BTFO

>> No.50443754

>expecting apple fags to know what benchmarks are

>> No.50443767

>mfw he thinks tiger vs leopard is cross platform
you gotta be a troll

>> No.50443785


Hey, we're not comparing snapdragon - we already know it's shit dickface. The OP was about IPHONE6S!

You asked for a site that compared across platforms, which I provided (and you ignored arstechnica). You also cherry pick out a fucking shitty snapdragon compared to a shitty AMD - WOW, BIG FUCKING DEAL. Now stop derailing with your red herring and accept I won.


Android poorfag can't match superior Apple.

>> No.50443796

>comparing an os to a company
Yeah, he's retarded guys. Go home

>> No.50443800


Look at the CPU architecture you bleeding double-nigger.

>> No.50443801
File: 27 KB, 273x200, Le-happy-merchant-orgasm-_54f8df42e4f8b61428d4af3f26a0f7eb.jpg [View same] [iqdb] [saucenao] [google] [report]

> all this Apple praise

>> No.50443822


You lost shitstain.

>> No.50443827

>Hey, we're not comparing snapdragon
That isn't the point, you are claiming geekbench can be used to compare across platform. I literally just proved you CAN'T do that. It doesn't matter if it's apple, snapdragon, intel, AMD, or fucking Exynos for fucks sake.

The simple fact geekbench scores an A4-5000 the same as SD600 when the A4-5000 is clearly superior debunks the fact that geekbench is a reliable benchmark tool across platform

I dont need to do anything else here buddy, you've proven yourself wrong with that link.

>> No.50443845

Bitch you were proven wrong multiple times by the other guy, so you tried to switch the subject to Android vs iOS. Fuck off lmao

>> No.50443861

iPhones are good phones. I reccommend them to my parents and people who dont give a shit about using their phone outside of social media. But I hate these applefags. Just deal with it. You "won" one benchmark after claiming for years they don't matter than the benchmark is shown to not be accurate at all. Just stay happy you're good at creating a polished simple ui and be happy lmao.

>> No.50443875


That always was the subject crater-face. Geekbench 2 was used in that particular benchmark and gives quite theoretical differences anyway. Superior Geekbench 3 should be more accurate

>> No.50443902

>more name calling and empty arguments trying to avoid the fact that you thought a mobile 3w chip is as powerful as a desktop 30w chip
Yeah, fuck off

>> No.50443913

Desktop =\= Mobile

It's apples and oranges trolltard

>> No.50443918

he's also still ignoring my post because he know there isn't any real excuse he can make up now that isn't just bullshit.

>> No.50443940

Wow its just like the lag hidden behind iOS animations.
>out of sight out of mind!

>> No.50443949

Apple no longer requires a $99 developers account to install unsigned apps. You will need Xcode and compile from source to get the apps on the iOS device.

>> No.50443950

Why can't you compare fruit? Are you retarded?

>> No.50443972

>Wow its just like the lag hidden behind iOS animations

>> No.50444045

Wow that icuck must be literally retarded or just burying his head in the sand to see this and not realize he is fucking wrong.

Ignorance is bliss as they say.

>> No.50444049


But it is bitchtits

>> No.50444071


Inferior Geekbench2. Superior Geekbench3 clears all that up. WINNING

>> No.50444136

What proof do you have of that?

You were asked to show a site that compared between platforms, you did and it showed that comparing across platform doesn't mean anything.

And all we have for proof geekbench 3 clears it all up is you claiming it does without any evidence, nor any real reason it WOULD clear it up. Its a fundamental problem of comparing different platforms.

Honestly at this point you're just gonna come out and say
>I was only trolling anyway guise

>> No.50444163

Ok. Prove to me that an iPhone 6s can run auto desk inventor with a browser next to it with the part you're trying to create while a music streaming service in the background while you have Skype on a second monitor to stay connected with your coworkers. Fuck off

>> No.50444180

This is UNDENIABLE PROOF that the Macbook is complete and utter shit.

>> No.50444207

No, its just proof OP and whoever OP grabbed that screenshot from have no understanding how benchmarks work.

Though macbooks are pretty shit.

>> No.50444290


>What proof do you have of that?
>And all we have for proof geekbench 3 clears it all up is you claiming it does without any evidence

Check the links in original post. You'll see superior accuracy comparing iPhone to a Mac.

>Honestly at this point you're just gonna come out and say
>I was only trolling anyway guise

What if I am? Would you feel a little silly?

>> No.50444309


Coming right up....

>> No.50444332

>superior accuracy comparing iPhone to a Mac
What the fuck is wrong with you? What accuracy? how can you claim that is accurate? An iphone is NOT as powerful as a macbook you fucking retard.

Jesus christ I seriously hope you are trolling. And it wouldn't make ME feel silly, if anything it would just make me wonder why someone would shill for apple this hard for free, seems like post purchase rationalization. Do you own an apple product m8 or are you just trying to stir up trouble on /g/ because you have nothing better to do with your life?

>> No.50444390

Newer chip slightly faster than older chip. Who would thunk it?

>> No.50444393

>Geekbench measures processor and memory performance only, and does not consider factors such as drive speed and GPU capabilities while calculating its score. While this limits the tool’s overall effectiveness in judging a device’s true potential, it’s focus on CPU and memory has enabled it to become truly cross platform, with versions for Windows, OS X, Linux, Android, and iOS. In theory, scores between devices should be directly comparable, allowing users to perform, for example, a comparison between the computing power of an iPhone and a 12-core workstation.

>Note that these lower numbers don’t mean that the device is getting slower; it merely means that, due to the changes in the latest version of Geekbench, results between versions 2 and 3 of the application should not be used for comparisons of performance.


Superior accuracy.

>> No.50444408

>seems like post purchase rationalization.
Wouldn't be surprised considering iPhone 6s + costs in excess of $1000. That money will get you a pretty nice laptop these days.

>> No.50444433


Jokes on you, I don't even own an iPhone.

>> No.50444505
File: 73 KB, 400x365, 1380232286324.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.50444507

>The intel CPU has a TDP of 35 watts
>the A9 (if its similar to previous SoCs from apple) will have a TDP around 3-4 watts.
Great to see Apple showing the world how shitty Intel and x86 really is.

>> No.50444513

>implying it actually does this all perfectly without error

Seriously though, you're just posting marketing bullshit there is no proof in any of that that what they are doing is actually accurate or directly comparable across platform like they claim.

Again, lets just go look at the geekbench 3 browser for some results.

Intel i5-3570 single core performance is ~3100

Iphone 6s single core performance is ~2400

So by this logic iphone 6s should only be around 30% slower than an intel i5-3570 in single core operations.

We KNOW this isn't true though. We KNOW a dekstop i5 is more than 4x as powerful in single core performance. So why the fuck does geekbench consider them so close?

The mental gymnastics you're performing here are shocking

>> No.50444552

>We KNOW a dekstop i5 is more than 4x as powerful in single core performance.
[citation needed]
You're fighting bullshit with bullshit.

>> No.50444594

Because I like having Kali Nethunter on my Nexus

>> No.50444633

The iPhone 6S can edit 2 streams of 4K video directly on the phone in iMovie. You bet your ass it's not that far off from Intel's laptop and desktop offerings.
Now we wait for the A9X figures on the iPad Pro.

>> No.50444680

>Granted we're comparing across platforms/browsers here, but the 5s as a platform does extremely well in Intel's favorite benchmark
>We're not talking about Haswell or even Ivy Bridge levels of desktop performance, but rather something close to mobile Core 2 Duo class

>> No.50444727
File: 169 KB, 1801x880, hurrrdurrr muh applels.jpg [View same] [iqdb] [saucenao] [google] [report]

I even used the i3-4330 for you to make it seem more fair. Both are dual core after all.

Even with the improvements made in the A9 chip for the 6s, it still wont be as powerful as a desktop CPU, just stop being retarded please.

>> No.50444777


Now do the comparison with a core 2 duo. It's a desktop cpu, right????

>> No.50444778

>browser benchmarks
Jesus christ, you intel shills are desperate.

>> No.50444813


>> No.50444833

>intel shills
I can't believe there are people actually trying to claim the iphone is even close to an intel CPU in performance, cause it just isn't.

Fuck it, lets break out an athalon 64x2 if we are going that far, not like that 90nm process node vs 20nm is gonna make any difference, or 10+ years of architecture refinements.

Not to mention, we are still ignoring the fact we are comparing x86 cores to fucking ARM cores.

This whole argument is fucking retarded, it isn't even a contest. An apple iphone is NOT as powerful as a desktop, and it isn't as powerful as a macbook from 2012-2013 like is insinuated in the OP.

comparing ARM to x86 in this manner is meaningless.

>> No.50444857

Android demonstratively uses moar cores.

Wonder if Apleplebel would post data on core usage within ishitOs

>> No.50444883

>Intel Core i3-4330, 3.5 GHz (2 cores)
>Geekbench score: 3126
>iPhone 6s, A9, 1.8 GHz (2 cores)
>Geekbench score: 2500

Seems to tally up fam

>> No.50444888
File: 1.69 MB, 1266x686, fug.png [View same] [iqdb] [saucenao] [google] [report]

Holy fuck the shilling is intense in this thread.

Is there a technology board on another site that isn't full of reddit and tumblr fags that don't know how to use a computer, and somehow think they can justify their expensive and terrible purchases?

>> No.50444892
File: 2.84 MB, 960x540, 1442536613143.webm [View same] [iqdb] [saucenao] [google] [report]

M8, you're wasting your breath.
He's an apple iCunt and will shill for apple to try and justify buying a $1200 phone that's just as fast as a $500 Android device or windows phone in real world use.

Synthetic benchmarks don't mean shit.

IPhone 6 has a faster cpu than the nexus 5, but the nexus 5 still runs faster despite losing in single thread performance.

>> No.50444908

Oh so the most recently created phone is the fastest?

>> No.50444920
File: 109 KB, 625x626, 1442794723997.jpg [View same] [iqdb] [saucenao] [google] [report]

>2ghz arm in iPhone is faster than 1ghz x86 in laptop

Please tell me you're not actually this retarded...

>> No.50444921

except in the kraken bench i posted where the i3 was literally almost 4x as fast...

Right, totally makes sense that something that is only 30% higher scoring on geekbench is 4x faster in kraken.

>Seems to tally up fam
maybe in your fucking head.

>> No.50444931


Innovative and beautiful too. We think you're gonna love it.

>> No.50444934

Double trips.

>> No.50444935

Not them but calling geekbench results bs and then trying to prove a point with browser benchmarks is not a good argument. Different browsers, different power saving methods, throttling and on. Leaving all fanfagging aside, A9 and A9X are seriously powerful chips if they can edit 4k streams as Apple claims. They won't be far off from your usual desktop performance if they can deliver that.

>we are comparing x86 cores to fucking ARM cores.
ARM servers are going to be a real thing soon. I would take the threat seriously if I were Intel.

>> No.50444955
File: 421 KB, 1680x1050, Certified Shit Wrecker.jpg [View same] [iqdb] [saucenao] [google] [report]

Hey apple-head the gay club´s two blocks down

I´ll show you what MUH CORES means once K12 is released.

Also, post the exynos 8890 benchmarks now

>> No.50444956

Don't all of these apps have to pull content off the network to display their main screen, completely invalidating the comparisons?

>> No.50444968

>Not them but calling geekbench results bs and then trying to prove a point with browser benchmarks is not a good argument

I wasn't using it to argue anything besides geekbench not being able to accuratly compare cross platform, I never pretended to think comparing browser benchmarks is ACTUALLY a decent way to compare CPUs, because it isn't And there is no EASY or GOOD way to compare ARM vs x86 in this situation, which is why no serious tech site does it.

>> No.50444976


>except in the kraken bench i posted where the i3 was literally almost 4x as fast
>differing browsers, optimizations etc.

You just can't accept the facts man. When comparing integer, floating point etc. the Geekbench scores are sound. You're mixing your results between different benchmarks with compounding variables.

Remember to take your meds.

>> No.50444982


But it doesn't look much different than the older version, and I didn't like the way that looked.

>> No.50444983
File: 153 KB, 750x750, 1442167841646.jpg [View same] [iqdb] [saucenao] [google] [report]

The double trips want an answer!

>> No.50444985

Until Microsoft moves to arm and mainstream developers start releasing big titles for arm I just don't see that happening any time soon.

>> No.50444992


>serious tech site does it

Anandtech, arstechnica .. etc.


>> No.50444998
File: 829 KB, 1440x2560, 2015-09-24 00.15.27.png [View same] [iqdb] [saucenao] [google] [report]

>Why haven't you upgraded to the latest and greatest /g/?

because why? this has a better camera, better battery life, and a better screen. also get more storage for less money.

also fuck ebay seller sobeonline1. niggers took my cash and didnt even bother sending the phone out.
>tfw it was the only store with the gold s6 in stock

>> No.50444999

OP is faggot

>> No.50445009

>completely invalidating the comparisons
How? unless you are insinuating android would somehow get network priority...? They are both connected to wifi, it doesn't invalidate anything at all.

>> No.50445024

I think he just means android has better internet speeds than iPhone :')

>> No.50445032

Server market mainly use FOSS stacks, they don't rely on microsoft software like desktop/workstation segment.

>> No.50445033

>When comparing integer, floating point etc. the Geekbench scores are sound

Seriously its closed source you moron, prove that it can actually be used to compare cross platform, fucking prove it you nigger.

>> No.50445066



You're welcome.

>Seriously its closed source you moron, prove that it can actually be used to compare cross platform, fucking prove it you nigger

You're the one claiming it cannot. I am the one claiming it can (like the developer and countless respected tech sites).

Therfore, you prove it cannot as you claim.

>> No.50445070

So the gimped macbook CPUs are so crippled, a shitty plebphone is now more powerful. Is this really news?

Literally every android application has a thread dedicated to the GUI, and only the GUI.
Anything mildly intensive is done in separate threads (unless you want the OS bitching at you permanently).

Using single-core benchmarks as an indicator of overall performance is like trying to guess how fast a car can go based on the size of the engine alone.

20 years ago that would have been enough I guess...

>> No.50445080
File: 23 KB, 680x367, iphone6sbenchmark-1.png [View same] [iqdb] [saucenao] [google] [report]

Jimmy left intel´s pandora box ready.
It´s scheduled to open on 2016.

This happened in 2005 too with powerpc cpu´s being more powerfull than x86 because both intel and amd floped

Also this is obligatory.

Also the stupid shit amounts of RAM and cores on ARM devices are now usable thanks since android can into 64 bit also this is not 2008, everything is optimized for atleast 4 cores

>> No.50445086

So glad my son isn't a faggot who cares about phones. Your fathers must be so ashamed of you plebs

>> No.50445088

>You're the one claiming it cannot

you literally cant prove ANYTHING because they wont show me the source code

If you give me the source code i'll fucking point out exactly why its a shitty comparison across platforms.

Fuck, the damn fact they DONT release the source code is probably because they know they are fucking with the numbers in some places.

>> No.50445175

The 2012 macbook air was seriously slow for its time.
Also it throttles
Also its shit
Also we are just comparing floating point performance

>> No.50445178


So, what you are saying is you cannot prove your claim that it is not good to compare cross platform because it is closed source? Why don't you trust closed source software anon? Why would they lie about results?

>> No.50445196

>links Gruber as a unbiased reason why ios is better

>> No.50445200

what I am saying is no one can prove anything one way or the other unless they release the source code, the only proof you have that it's accurate is they say it is.

>> No.50445211


>> No.50445213

what does being poor have to do with anything? an iphone and an android phone are the same price on contract

>> No.50445248

there is a reason the app costs $0.99 and $10 on windows.

They are a business anon.

>Why would they lie about results?
Because they want money.

>> No.50445271

>1.1 ghz

Oh my god, it's even better than my old Pentium 3 850 mhz! That's a chip that's only SIXTEEN YEARS OLD!

>> No.50445274

There are some great Android devices you can get on a budget too though.
For example:
>nexus 5
>zenfone 2
>xiaomi phones
>Lenovo k3 note
>doogee x5

That's the good thing about Android. You get low end, midrange, and high end devices.
On iPhone you only get shit tier devices.

>> No.50445284

>no one can prove anything one way or the other
Well that's not exactly true, certain some ones can do that. Deassembling and static analysis is a thing. There are guys who can lives a winning by analyzing machine code. But that's turned into a retarded discussion

>> No.50445307

>Because they want money.

Surely as a business, delivering accurate, reproducible results through a reliable app is more beneficial than a risky manoeuvre that has no gain and massive financial repercussions.

>> No.50445313

Wow that performance is insane. I'll be able to text, check my emails, make phone calls, and casually browse the internet faster than everyone else. Fucking plebs. When are they going upgrade to superior hardware like me?

>> No.50445326

good god, are you seriously doing this shit for free? This is getting impressive how far you're willing to shill.

>> No.50445333


Thanks fam.

>> No.50445348

>implying the average user needs anything more than a $49 smartphone


>> No.50445376
File: 24 KB, 630x686, 1442967082024.png [View same] [iqdb] [saucenao] [google] [report]

wasn't a compliment.

I'm going back to /a/ this place is getting normie as fuck lately.

>> No.50445386

Thank you, OP, for further turning /g/ from a technology board into a consumerist shitfest. Bravo, we need more threads like yours.

>> No.50445393


Anytime bro

>> No.50445424

>he actually stayed in the thread the entire 3 and a half hours

Seriously, you do this shit for FREE?

>> No.50445433

Apparently Geekbench doesn't use the same workloads on "mobile" as it does desktop: http://www.primatelabs.com/geekbench3/workloads.pdf

>> No.50445440
File: 384 KB, 350x263, 1442794362215.gif [View same] [iqdb] [saucenao] [google] [report]


>> No.50445469

Still no root access, unfortunately. They've also made jailbreaking a lot harder with rootless.

>> No.50445490
File: 71 KB, 768x1024, 1442883945866.jpg [View same] [iqdb] [saucenao] [google] [report]

Shush, let those retards think they made a smart choice and leave the Android market alone.

>> No.50445555

Well, Apple's cores are much better than whatever 8-core shit you can currently get in an Android phone. Maybe if Intel pushed a Core M-based phone that could change.

>> No.50445615
File: 1.51 MB, 250x250, 1442741814788.gif [View same] [iqdb] [saucenao] [google] [report]


>> No.50445665


Hey man, I got 5 kids to feed

>> No.50445667

The benchmarks show Apple's cores blowing other ARM processors out of the water, and Intel getting similar scores with ~4 times the workload.

>> No.50445709

Hey man, ah leave me alone you know
Hey man, well Henry, get off the phone, I gotta
Hey man, I gotta straighten my face
This mellow thighed chick just put my spine out of place

>> No.50445746

Of course, the totals are warped by the benchmarks that have special purpose acceleration on some SoCs, like crypto and possibly some of the compression/decompression benches. The Lua benchmark is probably the most interesting to look at.

>> No.50445755
File: 133 KB, 483x698, 1442886083956.jpg [View same] [iqdb] [saucenao] [google] [report]

If you say so.
Synthetic benchmarks don't mean shit against real world usage though.

>> No.50445756
File: 132 KB, 660x451, 264fbow.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.50445809

You mean like that nexus 5 vs iPhone 6 video people keep posting here? :')

>> No.50445821

Do you really believe that the "octacore" "2.5GHz" SoCs fare any better in the real world than Apple's and Intel's more balanced dual-cores? (That poor dinoasaur.)

>> No.50445842

People arguing about what botnetphone is better should literally kill themselves right now. Stop shitting up /g/ you fucking manchildren.

>> No.50445867

If it's about startup times, it probably comes down to shitty NAND.

Everyone knows that well-adjusted people have nothing to hide. And for the rest of us, we can avoid Gapps on Android phones.

>> No.50445915
File: 396 KB, 700x1050, 1402090316264.jpg [View same] [iqdb] [saucenao] [google] [report]

>Poorfags need not apply.
Says the person who buys a phone on a 2 year plan ? That gets jewed with a 16gb base model.

>> No.50445947

Yes, because >>50445070

Do your research, more cores has an advantage when the os and it's apps are able to take advantage of them.

Your single core performance is 2% faster and your device is limited to 2 cores.

Synthetic benchmarks don't mean shit, it's real world usage that matters.

>> No.50446012

Since cpu power isn't even that important anymore I'd take an a6 7400k with an ssd over a g3260 with a 5400rpm hdd any day.

Same goes for phones. If the nexus 5 has faster nand than iPhone and things load faster then why would I get an iPhone?

Does the average user even do anything that a pentium 4 wouldn't be able to handle?

>> No.50446051

>but single-core performance is a better measure for the sort of things you can feel while using a device.
What a fucking joke.

>> No.50446078

>They are both connected to wifi

proof? And proof that the wifi router is not sabotaging the iphone?

>> No.50446105

I know right.
If my xperia x10 had more ram I'd probably still be using it.

I still have my core 2 duo at home e8500, the ipc is horrible on such an old processor but it's still giving no signs of lag for word processing and internet browsing on mint.

Single thread performance is starting to become less relevant these days.

But shills gonna shill, lel

>> No.50446118

Has any site posted the full output of the benchmark or do they all only post the number?

Cuz, you know, there's a lot more to the result of a geekbench benchmark than just the number it spits out at the end. It gives you all the results of the individual tests it runs and it would be interesting to know if it is getting great results across the board or in specific areas.

>> No.50446126

Yeah, because it's so useful that the scheduler puts that IO thread on its own core instead of timesharing it on the already active one. And of course no CPU intensive apps have sequential parts that the rest of threads have to wait for, everything just scales perfectly to n processors.

Apple has been ridiculously stingy with both RAM and flash, but you have to admit that their SoCs are top notch. For me the old ARMv7 in my phone is more than fast enough, but if it was faster I could probably get better battery due to race to sleep (assuming non-shitty software, that is).

>> No.50446132

look at the webm faggot, both have full bars for wifi signal strength.

Also, how the fuck could it POSSIBly sabotage the iphone to load, but specifically load SLOWER than the android phone?

You're just being obtuse.

>> No.50446212

Their SoCs are good, but in real world usage is not going to make a difference.

If apple went quad core, put in a bigger battery, bumped up to 2gb ram(think they did that for the 6s plus?) then they would be significantly better than most android devices.

>> No.50446265

But then they would have to make their phones thicker!

>> No.50446297
File: 47 KB, 184x184, tumblr_n6obizsDME1rpwm80o1_250.jpg [View same] [iqdb] [saucenao] [google] [report]

>800+ dollars for a phone that you can't open to change battery or sd card, will crack it's own screen if you look at it the wrong way, has to be traded in and upgraded for a new version every 3-4 for a 200 dollar fee

get this shit outta here nigga. unless you are a white teenage girl or a superstar there is no reason to use apple ever

>> No.50446365

Both models have 2 GiB RAM. I don't see why you would add more cores in a consumer device unless you're marketing the core count to tech illiterates/"power user"; 3 or 4 GiB would probably be more useful, at least at the end of the device's lifespan. SMT could also probably squeeze out some performance without duplicating all the logic.

>> No.50446426
File: 91 KB, 671x610, geek.png [View same] [iqdb] [saucenao] [google] [report]

>try out geekbench on my PC
>this is my cpu utilization
Is this a joke? It never used the full capability of every core. Could it have been memory starved or is it just shit?

>> No.50446459

Phone touchscreens are inaccurate and poorly done.

Phone OSes are underpowered, underfeatured, and contain spyware.

Phone applications are 95% redone web pages and are mostly spyware.

Phones track you via the cellular network and is anti-privacy.

Phones do nothing you can't do with a little preparation and common sense before you leave to go out.

Phones are shit and must die right the fuck now. You dumb goddamn little children have fallen for this goddamn viral marketing push to put a fucking tracking device in everyone's goddamn pocket and you are actually willing to pay for the "priviledge".

You fucking stupid little children. God fucking damn you all and god fucking damn companies and governments working together to usher in this new era of a neutered, monitored soceity of idiots. You could have stopped it, and if you hurry and come to your fucking senses before this shit becomes mandatory or they do something much worse, you can still stop it.

Throw your shitphone away.

>> No.50446471

>Phone touchscreens are inaccurate and poorly done.
Swing and a miss.

>> No.50446537

All benchmark scores are dick waving.

>> No.50446543

>Swing and a miss
Just like every single attempt to use a fucking touchscreen.

They're terrible and you're terrible for supporting this ANTI-TECHNOLOGY garbage, child. Go fuck yourself if you can't understand this SIMPLE idea.

>> No.50446555

>Phone touchscreens are inaccurate
Stopped reading there. Lose weight, fatty

>> No.50446566

Someone's pretty mad about having piss poor motor skills.

>> No.50446568

SPECint2006 or death.

>> No.50446584

And it took nearly 200 posts before anyone mentioned SPECint.

/g/ is sad

>> No.50446613

>utterly disgusting, look at all those cringy whites, awful motion/interface/gesture design
>slower now than lagdroid

meanwhile in android
>easily get to lollipop then to marshmallow if you actually know what you're buying
>material design

>> No.50446621

You're an idiot. UI elements smaller than even a goddamn child's finger having to be touched accurately IS NOT FUCKING ACCURATE, YOU DUMB GODDAMN LITTLE SHIT.

Hell, you're probably a goddamn shill for one phone company or the other (or worse, some government shit that's trying to keep their monitoring devices out there) because no one is this stupid without being paid to be.

>> No.50446660

And by the way, PAID SHILLS, I love how you harp on one little thing you can troll about instead of the much more damning evidence against your shitphones afterwards.

Go fuck yourselves. Your little tracking device toys are dead and done. No one needs one and no one should want one.

>> No.50446673

Bad UI design is not a fault if the technology you shitcunt.

You're probably one of those faggots that changes their phone DPI so far that apps believe they're running on a fucking tablet and use those UI layouts.

You're the problem.

>> No.50446825

So should I go from Note 4 to iPhone 6s?

>> No.50446854

Security risk

>> No.50446911


>> No.50446982

Idk. I really don't care about brand. I just want a kick as phone with a big screen.i liked the Note 4. Was thinking about getting the 5 but a m8 said I should get a 6s.

>> No.50447075

But what's wrong with your 4?

Why do you need to upgrade? Is it slow?

>> No.50447085

it was mentioned several times earlier, you just had to have been there man

>> No.50447164

That's like buying a mansion only to have 2 rooms unlocked because someone might enter rest.

>> No.50447432

Not at all m8. I am due for an upgrade with my contract. I guess since it's paid off I can just use it and my bill will be ~35 dollars cheaper.

>> No.50447565

That might be best, put the money somewhere it's needed more.

>> No.50447680

Yeah I agree with you. I've always gotten upgrades whenever I was eligible so I guess I was used to getting something new. But my Note 4 is the best phone I've had. I guess there is no reason to get rid of it.

>> No.50447793

Well he linked daringfireball in th OP

>> No.50447879

>pays 1500 dollars for a phone and is very salty about his subpar purchasing decision.

>> No.50447927

Apple products are for homosexual men.

>> No.50448395

Now this anon knows what he's talking about.

>> No.50448627

WTF do you think a processor does? It processes information as fast as possible. What do you think a benchmark does? It tests to see how fast a processor can process the information.

For fuck's sake, are you this retarded?

>> No.50448674

It's not like there's much of a choice when all of their competitors ship Shitdroid on their devices.

>> No.50448741

Ok, see how processing is done is very interesting. Modern ARM and x64 processors have dedicated instructions for encryption & decryption, cryptographically secure hashes and so on. Now that you're aware:

Geekbench3 is a flaming piece of shit that's biased for a few reasons: most common scores for Windows are x86 (which is slower than x64 because it doesn't have registers r8 to r15. This benchmark doesn't use x64's AES/SHA2 instructions for their corresponding benchmarks while it does for ARMv7 and ARMv8. This is a large component in the integer performance score. In result, comparing x86 chips to Apple's with Geekbench is basically broken. Use SPECint like everyone else in the industry or shut the fuck up.

>> No.50448952

Daily iphone vs android: The thread

>> No.50448954

does x86 even have access to native 64 bit integers? or is handling 64 bit integers in general faster on x64

>> No.50449001


quit worrying about the x86 comparison... it's about the ARM chips inside of the new iPhone and new flagship Samsungs... everything else, including the drivel about comparing to laptop/desktop processors is just bait.

The fact that Apple engineers can create a in house, dual core ARM processor that out performs an eight core processor Samsung throws into their phones is pretty amazing.

>> No.50449035

8 core android phones don't use all 8 cores at the same time you dumb fuck
it has 4 low power cores for basic every day things, and 4 high power cores for intensive things (like benchmarks)
all 8 cores are never active at the same time

>> No.50449077

>wanting shItOS on a phone
you'd might as well have a dumb phone

>> No.50449691

There's a reason why it's called a big little configuration.

>> No.50449770


>> No.50450558

How do SoCs like the Snapdragon 808 with fewer big cores than small cores work then, smartypants?

>> No.50450589


Sync them with VLC, why the fuck do you need a file browser outside of the app?

>> No.50450631


Synthetic browser benchmarks are total trash for determining CPU performance, especially when you're comparing two different CPU platforms using entirely different browsers.

Geekbench is at least a level playing field when it comes to benchmarking different architectures.

>> No.50450646


Pretty much.
And it seems the Android has the apps pre-cached (tombstoned) too, while the iOS apps are loading from scratch.

>> No.50450710

>Geekbench is at least a level playing field when it comes to benchmarking different architectures.
Geekbench 3 uses different datasets for mobile and desktop platforms, see: >>50445433

Also, if want to extrapolate to general performance, the accelerated subtests skew the totals.

>> No.50450883
File: 102 KB, 229x215, 1442196622289.png [View same] [iqdb] [saucenao] [google] [report]

>work at ups
>have to start 15 minutes early tomorrow because of fucking 6s shipments
I hope everyone who owns any apple products dies a horrible death

>> No.50451033

>15 minutes early
Is that all? I have to start 3 hours early. Stop being a whiny cunt.

>> No.50451101

Do you know that modern CPUs are typically turning some cores off when there is no need for them and lowering their frequency?

For the OS it doesn't matter if SoC is lowering frequency and disabling cores, or if there are two logical blocks of CPUs (in this case 4 high performance cores and 2 low performance).
It will use what is available in the moment.

>> No.50451117

>Apple is literally years ahead of the industry.
How to spot a paid review or a religious shill.

>> No.50451137
File: 12 KB, 497x203, filter.png [View same] [iqdb] [saucenao] [google] [report]

Reminder to ignore iShill threads.

>> No.50451140

steal a 6s for me pls

>> No.50451141


Someone post the cats vs dogs benchmark

>> No.50451370

>Yeah, because it's so useful that the scheduler puts that IO thread on its own core instead of timesharing it on the already active one.

Whether it's a good idea or not to force everything to run that way is another argument, but in real world usage android benefits greatly from having at least 2 cores.
Time-sharing threads on the same core is stupid, there is no point having threads at all if that's what you plan on doing.

>And of course no CPU intensive apps have sequential parts that the rest of threads have to wait for, everything just scales perfectly to n processors.

Of course the threads rely on each other, and the performance doesn't scale perfectly with MOAR COARS. But it does scale very well from 1 core up to 2, 3 or 4 cores.

>> No.50451538

Antutu says otherwise...

>> No.50451585

Samsung got flak for "doping" their benchmarks around the Note 3, particularly in Geekbench, by forcing the CPU to run at max clock and not letting it idle.
But it wasn't overclocking the CPU, 2.3GHz was the max stock clock, but then why did it make a 20% difference in scores?
Wouldn't that mean that under normal conditions it isn't loading the CPU to the point where it would clock up to its maximum clock speed?
Wouldn't that make score comparisons difficult because you need to manage what percentage of the maximum clock the CPU is actually getting to and then weigh that with what other CPUs get to and their scores and all that shit?

>> No.50451856
File: 185 KB, 1000x1000, 29940346_jpg - (184_95KB, 1000x1000).jpg [View same] [iqdb] [saucenao] [google] [report]

>Operating systems can become bankrupt.

>> No.50451915

Mobile SoCs are not designed to run at their maximum speed for long periods of time.
There are even bigger limitation in terms of maximum power consumption and heat output on phones and tablets than on notebooks and desktops.
Running at maximum speed is, for most SoC, similar to turbo on desktop CPUs.
It will run on those frequencies and on as many cores as long as possible.

>> No.50452070

But you can run turbo frequencies on your desktop CPU so long as the temps stay in acceptable range.
I downloaded a Perfmon app and ran it while running Geekbench and none of my phones CPU cores reached 100% utilization and the CPU clock speeds, while were often at their max clock, would often fluctuate with no pattern toward single and multi threaded operations. Thermals weren't an issue within the one minute required to run the test either.
My problem is this, if the benchmark does not stress the CPU to its limit then how can you really compare the results when it is basically up to the CPU (or the OS) to decide when it is going to clock up?

Name (leave empty)
Comment (leave empty)
Password [?]Password used for file deletion.